Multi-modal Question Answering on MMBench-CC
0.645AccuracyVisionThink-7B
Evaluation results
| Method | Links | |
|---|---|---|
| VisionThink-7B2025.10 | 0.645 | |
| MergeMixbaseline=SFT Vision2025.10 | 0.6431 | |
| SFT Vision2025.10 | 0.6352 | |
| Qwen2.5-VL-Ins-7B2025.10 | 0.6294 |
| Method | Links | |
|---|---|---|
| VisionThink-7B2025.10 | 0.645 | |
| MergeMixbaseline=SFT Vision2025.10 | 0.6431 | |
| SFT Vision2025.10 | 0.6352 | |
| Qwen2.5-VL-Ins-7B2025.10 | 0.6294 |