Multimodal Understanding on MMBench Chinese (test)
63.6AccuracyLLaVA-1.5
Evaluation results
| Method | Links | |
|---|---|---|
| LLaVA-1.5Language Model=Vicuna-13B2023.11 | 63.6 | |
| ShareGPT4V-7BLanguage Model=Vicuna-7B2023.11 | 62.2 | |
| LLaVA-1.5Language Model=Vicuna-7B2023.11 | 58.3 | |
| Qwen-VL-ChatLanguage Model=Qwen-7B2023.11 | 56.7 | |
| IDEFICS-80BLanguage Model=LLaMA-65B2023.11 | 38.1 | |
| InstructBLIPLanguage Model=Vicuna-7B2023.11 | 23.7 | |
| LLaVALanguage Model=Vicuna-7B2023.11 | 14.1 | |
| Qwen-VLLanguage Model=Qwen-7B2023.11 | 7.4 |