Generative Question Answering on NYT (test)
76.42EMSWEP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SWEPModel Architecture=T5-small2021.05 | 76.42 | 84.81 | |
| Prior-AugModel Architecture=T5-small2021.05 | 75.52 | 84.04 | |
| MLEModel Architecture=T5-small2021.05 | 75.1 | 83.88 | |
| Gaussian-DropoutModel Architecture=T5-small2021.05 | 74.76 | 83.57 | |
| Word-DropoutModel Architecture=T5-small2021.05 | 74.27 | 83.24 | |
| Bernoulli-DropoutModel Architecture=T5-small2021.05 | 73.81 | 83.06 | |
| Adv-AugModel Architecture=T5-small2021.05 | 73.46 | 82.84 | |
| SSMBAModel Architecture=T5-small2021.05 | 73.29 | 82.79 |