Mathematical Reasoning on GSM8k Tensor-2 (test)
49.2AccuracyAdaBack
Evaluation Results
| Method | Links | |
|---|---|---|
| AdaBackModel Size=3B2025.06 | 49.2 | |
| SFT+RLModel Size=3B2025.06 | 42.7 | |
| SFT+AdaBackModel Size=3B2025.06 | 42.2 | |
| SFT+AdaBackModel Size=1B2025.06 | 11.3 | |
| AdaBackModel Size=1B2025.06 | 8.5 | |
| SFT+RLModel Size=1B2025.06 | 6.9 | |
| Base+RLModel Size=1B2025.06 | 0 | |
| Base+RLModel Size=3B2025.06 | 0 |