Comprehensive Multi-image on MuirBench
62.3AccuracyGPT-4V
Evaluation results
| Method | Links | |
|---|---|---|
| GPT-4VVersion=V-Preview2024.08 | 62.3 | |
| LLaVA-OneVision-72BModel Scale=72B2024.08 | 54.8 | |
| LLaVA-OneVision-7BModel Scale=7B2024.08 | 41.5 | |
| LLaVA-OneVision-0.5BModel Scale=0.5B2024.08 | 23.5 |