Tool-use reasoning on ToolAlpaca
66.73AccuracyMOPD
Evaluation Results
| Method | Links | |
|---|---|---|
| MOPDbackbone=Qwen3-8B2026.05 | 66.73 | |
| SDPObackbone=Qwen3-8B2026.05 | 63.45 | |
| MCPO-GRPOBackbone=Qwen3-8B2026.05 | 61.45 | |
| MCPO-DAPOBackbone=Qwen3-8B2026.05 | 61.14 | |
| GRPObackbone=Qwen3-8B2026.05 | 60.32 | |
| MCPO-GRPOBackbone=Qwen3-4B2026.05 | 60.29 | |
| MCPO-DAPOBackbone=Qwen3-4B2026.05 | 59.87 | |
| MT-GRPOBackbone=Qwen3-8B2026.05 | 59.4 | |
| Qwen3-8B2026.05 | 59.11 | |
| MGSBackbone=Qwen3-8B2026.05 | 58.82 | |
| MT-GRPOBackbone=Qwen3-4B2026.05 | 58.67 | |
| MGSBackbone=Qwen3-4B2026.05 | 58.35 | |
| CLIPOBackbone=Qwen3-8B2026.05 | 57.69 | |
| GRPOBackbone=Qwen3-8B2026.05 | 57.32 | |
| GRPOBackbone=Qwen3-4B2026.05 | 56.73 | |
| DAPOBackbone=Qwen3-8B2026.05 | 56.42 | |
| CLIPOBackbone=Qwen3-4B2026.05 | 55.56 | |
| DAPOBackbone=Qwen3-4B2026.05 | 55.29 | |
| Base ModelBackbone=Qwen3-8B2026.05 | 47.35 | |
| Base ModelBackbone=Qwen3-4B2026.05 | 45.83 |