Zero-shot Language Understanding on MMLU, GSM8k, HellaSwag, and WinoGrande
72.8MMLU
FP16
Updated 1mo ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2025.09 | 72.8 | 85.1 | 80 | 78.9 | 79.7 | - | |
2025.09 | 70.64 | 82.26 | 79.24 | 77.35 | 77.37 | 98.02 | |
2025.09 | 70.57 | 82.71 | 79.3 | 77.03 | 77.4 | 98.06 | |
2025.09 | 70.52 | 82.49 | 79.35 | 76.95 | 77.33 | 97.96 | |
2025.09 | 70.27 | 82.56 | 79.18 | 76.64 | 77.16 | 97.76 | |
2025.09 | 70.25 | 80.52 | 79.01 | 76.64 | 76.6 | 97.05 | |
2025.09 | 69.38 | 81.8 | 79.41 | 77.9 | 77.12 | 97.71 | |
2025.09 | 69.26 | 80.82 | 78.52 | 77.03 | 76.41 | 96.8 | |
2025.09 | 68.79 | 81.43 | 78.4 | 76.88 | 76.37 | 96.76 | |
2025.09 | 68.23 | 80.36 | 77.26 | 75.93 | 75.44 | 95.58 | |
2025.09 | 68.16 | 78.7 | 78.56 | 75.3 | 75.18 | 95.25 | |
2025.09 | 66.24 | 77.56 | 77.34 | 74.11 | 73.81 | 93.51 |