Dockerfile

FROM ubuntu:22.04

ENV DEBIAN_FRONTEND=noninteractive

# Update and install necessary dependencies
RUN apt update && \
    apt install --no-install-recommends -y \
        build-essential \
        python3 \
        python3-pip \
        wget \
        curl \
        git \
        cmake \
        zlib1g-dev \
        libblas-dev && \
    apt clean && \
    rm -rf /var/lib/apt/lists/*

WORKDIR /app

RUN wget https://huggingface.co/itsdotscience/Magicoder-S-DS-6.7B-GGUF/resolve/main/Magicoder-S-DS-6.7B_q8_0.gguf

RUN git clone https://github.com/ggerganov/llama.cpp.git && \
    cd llama.cpp && \
    git submodule init && \
    git submodule update && \
    make

# Create a non-root user for security reasons
RUN useradd -m -u 1000 user && \
    mkdir -p /home/user/app && \
    cp /app/Magicoder-S-DS-6.7B_q8_0.gguf /home/user/app

RUN chown user:user /home/user/app/Magicoder-S-DS-6.7B_q8_0.gguf

USER user
ENV HOME=/home/user

WORKDIR $HOME/app

# Expose the port
EXPOSE 8080

# Start the llava-server with models
CMD ["/app/llama.cpp/server", "--model", "Magicoder-S-DS-6.7B_q8_0.gguf", "--threads", "12", "--host", "0.0.0.0"]