Hello. Recently I used the lmms-eval repo to evaluate the [checkpoint](https://huggingface.co/liuhaotian/llava-v1.6-mistral-7b). However, the evalution results are different from here [link](https://github.com/haotian-liu/LLaVA/blob/main/docs/MODEL_ZOO.md). <img width="841" height="255" alt="Image" src="https://github.com/user-attachments/assets/992b9ace-bb56-4b78-bcf9-1ba8ebc77f9d" /> <img width="861" height="256" alt="Image" src="https://github.com/user-attachments/assets/70548ea2-3811-414d-80bf-516ebfc7d8a6" /> <img width="871" height="214" alt="Image" src="https://github.com/user-attachments/assets/f6984971-0be3-4f07-99f6-44a94fba493e" />