Tool Usage on BigCodeBench (Pass@1)
41.67Pass@1Qwen2.5-7B-Ins + ADR
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-7B-Ins + ADRBase Model=Qwen2.5-7B-Instruct, Data Augmentation=ADR2026.05 | 41.67 | |
| Qwen2.5-7B-Ins + KodCodeBase Model=Qwen2.5-7B-Instruct, Data Augmentation=KodCode2026.05 | 41.27 | |
| Qwen2.5-7B-InsBase Model=Qwen2.5-7B-Instruct2026.05 | 38.3 |