---
license: llama2
---

The original model is here: https://huggingface.co/Lin-Chen/ShareGPT4V-7B  
This is the K-type quantized variant (for inference with llama.cpp llava-cli)  

In my tests this is currently the best llava based vision model