Liyan

遇到同样的问题+1

Yes I think I do have some trained weights for the base version. If I can find them I will give the link in README.

There might be sth off in your configuration. I have trained spanbert_base weight with F1 score 77+.

c2f 2019 is train_bert_xxx_ml0_d2. d2 uses Attended Antecedent as higher-order inference while d1 only uses local decisions.

@L-hongbin I was talking about spanbert_base, not bert_base

@L-hongbin The results are similar to the reported numbers but I don't have the exact numbers on my hand now.