thunlp / FewRel

A Large-Scale Few-Shot Relation Extraction Dataset
https://thunlp.github.io/fewrel.html
MIT License
731 stars 165 forks source link

Is there a discrepancy between paper and repository? #46

Closed wjczf123 closed 4 years ago

wjczf123 commented 4 years ago

BERT-PAIR for Domain Adaptation (FewRel 2.0) achieves 67.41 for 5-way 1-shot in the repository and website. But in your paper BERT-PAIR achieves 56.25 for 5-way 1-shot in your paper. Is there a discrepancy between paper and repository?

gaotianyu1350 commented 4 years ago

Hi, the results in the repo and the benchmark website are updated after further tuning the hyper-parameters. The method is exactly the same.

wjczf123 commented 4 years ago

OK. Thank you a lot.