xiangrongzeng / copy_re

Release for acl18 paper "Extracting Relational Facts by an End-to-End Neural Model with Copy Mechanism"

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

f1/precision/recall

LifangD opened this issue · comments

hi, Dr Zeng, when I use the default setting to train on webnlg dataset, the f1/precision/recall seems to be 0.00.. and the entity in visualized result repeated twice. still can't figure out what's wrong and looking forward to your reply~

Try learning rate 0.001 and use pre-trained word embedding

Thanks, it's exactly the problem of the learning rate.