Multilingual Multiple Choice Question Answering on HellaSwag
54.6Accuracy (EN)Standard Evaluation
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Standard EvaluationScale=8B2026.06 | 54.6 | 53 | 1.6 | |
| ParaEvalScale=8B2026.06 | 51.6 | 51.1 | 0.5 | |
| Standard EvaluationScale=3B2026.06 | 49.4 | 48.1 | 1.3 | |
| ParaEvalScale=3B2026.06 | 46.4 | 46.1 | 0.3 | |
| Standard EvaluationScale=1B2026.06 | 44.39 | 43.57 | 0.82 | |
| ParaEvalScale=1B2026.06 | 41.33 | 41.49 | 0.16 |