Question Answering on QASC-IR (test)
89.6AccuracyUnifiedQA 11B
Evaluation results
| Method | Links | |
|---|---|---|
| UnifiedQA 11BModel=UnifiedQA 11B2022.10 | 89.6 | |
| TEAMBackbone=DeBERTa Large2022.10 | 89.35 | |
| ScoreBackbone=DeBERTa Large2022.10 | 85.65 | |
| TEAMBackbone=RoBERTa Large2022.10 | 77.93 | |
| ScoreBackbone=RoBERTa Large2022.10 | 73.26 |