Question Answering on BoolQ (dev)
86.1AccuracyFine-tuning
Evaluation results
| Method | Links | |
|---|---|---|
| Fine-tuningTraining Setting=Full training dataset, Backbone=RoBERTa-large2021.04 | 86.1 | |
| EFLTraining Setting=Full training dataset, Backbone=RoBERTa-large2021.04 | 86 | |
| EFLTraining Setting=Few-shot with K=8, Backbone=RoBERTa-large2021.04 | 73.9 | |
| LM-BFFTraining Setting=Few-shot with K=8, Backbone=RoBERTa-large2021.04 | 71.2 | |
| MajorityTraining Setting=Full training dataset, Backbone=RoBERTa-large2021.04 | 62.3 | |
| Fine-tuningTraining Setting=Few-shot with K=8, Backbone=RoBERTa-large2021.04 | 60.8 |