Commonsense Reasoning on Hellaswag (fewshot=10, test)
88.57AccuracyPhiMoE (FP16 baseline)
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PhiMoE (FP16 baseline)Precision=FP16, Shot count=10-shot2026.04 | 88.57 | — | — | — | |
| NPUMoEPrecision=FP16, Shot count=10-shot2026.04 | 87.62 | 10.63 | 34.77 | -1.07 |