Multi-choice Question Answering on MMBench (test dev)
86.4AccuracyQwen2 VL
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2 VLModel Scale=7B, Model Modality Type=Vision-language2024.10 | 86.4 | |
| GPT-4oModel Modality Type=Proprietary2024.10 | 83.4 | |
| MiniCPM-Llama3-V 2.5Model Scale=8B, Model Modality Type=Vision-language2024.10 | 76.7 | |
| Baichuan-omniModel Scale=7B, Model Modality Type=Omni-modal2024.10 | 76.2 | |
| VITAModel Scale=8x7B, Model Modality Type=Omni-modal2024.10 | 74.7 |