Multimodal Understanding on VisWiz
58.4AccuracyStableLLaVA
Evaluation Results
| Method | Links | |
|---|---|---|
| StableLLaVAnumber of parameters=13B2023.08 | 58.4 | |
| LLaVA-1.5number of parameters=13B2023.08 | 53.6 | |
| Qwen-VL-Chat2023.08 | 38.9 | |
| IDEFICS 80Bnumber of parameters=80B2023.08 | 36 | |
| IDEFICS-9Bnumber of parameters=9B2023.08 | 35.5 | |
| Qwen-VL2023.08 | 35.2 | |
| InstructBLIP2023.08 | 33.4 | |
| BLIP22023.08 | 19.6 |