| Benchmark | Model | Rank | Results |
|---|---|---|---|
| visual-question-answering-vqa-on-mmvp | ESC + Qwen2-VL-7B | #5 | Paired Accuracy: 45.33 |
| visual-question-answering-vqa-on-realworldqa | ESC + Qwen2-VL-7B | #87 | Accuracy: 54.51 |
| visual-question-answering-on-mm-vet | ESC + Qwen2-VL-7B | #101 | GPT-4 score: 40.05 |
| multimodal-reasoning-on-mmmu | ESC + Qwen2-VL-7B | #126 | Accuracy: 40.02 |