liuhaotian

/

llava-v1.5-7b

like 359

I believe llama cpp uses the base model and the projector not them combined. So you should get a gguf of vicuna 1.5 and then get the projector and use the llama cpp script for multi model.