Supervised Fine-Tuning on Heterogeneous SFT Benchmarks (Evaluation)
65GSM8K Score
EPI
Updated 3d ago
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
2026.04 | 65 | 33.2 | 69.1 | 8.9 | 9.1 | 8.57 | |
2026.04 | 63.7 | 32.6 | 67.2 | 8.7 | 8.9 | 8.23 | |
2026.04 | 62.2 | 31.5 | 66.4 | 8.6 | 8.8 | 8.21 | |
2026.04 | 61.8 | 31.2 | 65.4 | 8.4 | 8.6 | 7.98 | |
2026.04 | 61 | 30.8 | 64.9 | 8.3 | 8.6 | 7.93 | |
2026.04 | 60.4 | 30.6 | 65.2 | 8.2 | 8.4 | 7.97 | |
2026.04 | 59.8 | 29.8 | 64.6 | 8.5 | 8.7 | 7.9 | |
2026.04 | 59.3 | 29.5 | 63.1 | 8.2 | 8.4 | 7.74 | |
2026.04 | 59.2 | 29.4 | 63 | 8.1 | 8.3 | 7.68 | |
2026.04 | 59.1 | 30 | 63.7 | 7.9 | 8.2 | 7.71 | |
2026.04 | 58.5 | 29.2 | 63 | 8.2 | 8.4 | 7.67 | |
2026.04 | 58.5 | 30.1 | 64 | 8.3 | 8.5 | 7.78 | |
2026.04 | 57.4 | 28.7 | 62.1 | 7.7 | 8 | 7.48 | |
2026.04 | 57.2 | 29.6 | 62.4 | 8 | 8.3 | 7.55 | |
2026.04 | 56.9 | 27.9 | 61.5 | 8 | 8.2 | 7.44 | |
2026.04 | 56.8 | 27.9 | 61.4 | 7.8 | 8.1 | 7.44 | |
2026.04 | 55.6 | 28.3 | 60.8 | 7.8 | 8.1 | 7.32 | |
2026.04 | 55 | 27.1 | 60.2 | 7.4 | 7.8 | 7.21 | |
2026.04 | 54.4 | 26.3 | 59.6 | 7.7 | 8 | 7.17 | |
2026.04 | 53.1 | 26.7 | 58.9 | 7.5 | 7.9 | 7.05 |