Math on OlpBench
49.1AccuracyQwen2.5-72B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-72BParameters=72B2026.05 | 49.1 | |
| Qwen2.5-32BParameters=32B2026.05 | 48.8 | |
| Qwen2.5-14BParameters=14B2026.05 | 45.1 | |
| FORGE-7B-NPParameters=7B, Configuration=NP2026.05 | 32.1 | |
| FORGE-7B-MATHParameters=7B, Configuration=MATH2026.05 | 31.8 | |
| FORGE-7B-NP-MATHParameters=7B, Configuration=NP-MATH2026.05 | 31.2 | |
| Qwen2.5-7BParameters=7B2026.05 | 30 | |
| Qwen2.5-3BParameters=3B2026.05 | 29.7 | |
| InternLM3-8BParameters=8B2026.05 | 25.1 | |
| Qwen2.5-7B-SFTParameters=7B, Configuration=SFT2026.05 | 21.4 | |
| LLama3.1-8bParameters=8B2026.05 | 17.7 |