FROM python:3.11.6-bullseye

ENV LLAMA_FAST=true

RUN apt update && apt install -y libopenblas-dev
RUN git clone --recurse-submodules https://github.com/spuuntries/llava-cpp-server
WORKDIR llava-cpp-server

WORKDIR llama.cpp
RUN git pull origin master
WORKDIR ..

RUN wget https://huggingface.co/mys/ggml_llava-v1.5-7b/resolve/main/mmproj-model-f16.gguf
RUN wget https://huggingface.co/mys/ggml_llava-v1.5-7b/resolve/main/ggml-model-q4_k.gguf

RUN make

CMD ["bin/llava-server", "-m", "ggml-model-q4_k.gguf", "--mmproj", "mmproj-model-f16.gguf", "--port", "7860", "--host", "0.0.0.0", "--log-http"]