

Q1: How to train GR-models? Or how is the model like ? 
    Please see the folder ./GR_models and the read_me file inside. 


Q2: How to apply the trained GR-model in LS-task ?
    Please see the folder ./LS_infer/ and the read_me file inside 

Q3:  How to evaluate the trained GR-model in unsupervised STS tasks ?
    Please see the folder ./STS_infer/ and the read_me file inside 

Q4: How to fine-tune the GR-model in supervised STS-Benchmark task ?
    Please see the folder ./STS-B_infer/ and the read_me file inside 