Reading Comprehension On Adversarialqa
평가 지표
D(BERT): F1
D(BiDAF): F1
D(RoBERTa): F1
Overall: F1
평가 결과
이 벤치마크에서 각 모델의 성능 결과
모델 이름 | D(BERT): F1 | D(BiDAF): F1 | D(RoBERTa): F1 | Overall: F1 | Paper Title | Repository |
|---|---|---|---|---|---|---|
| BERT-Large | 62.4 | 71.3 | 54.4 | 62.7 | Beat the AI: Investigating Adversarial Human Annotation for Reading Comprehension | |
| BiDAF | 30.2 | 28.6 | 26.7 | 28.5 | Beat the AI: Investigating Adversarial Human Annotation for Reading Comprehension | |
| RoBERTa-Large | 65.5 | 74.1 | 53.4 | 64.4 | Beat the AI: Investigating Adversarial Human Annotation for Reading Comprehension |
0 of 3 row(s) selected.