Dockerfile

FROM docker.io/library/ubuntu:22.04 AS builder
ENV DEBIAN_FRONTEND=noninteractive
ENV TZ=UTC
RUN apt-get update && apt-get install -y --no-install-recommends \
    build-essential cmake git libopenblas-dev curl ca-certificates \
    && rm -rf /var/lib/apt/lists/*
WORKDIR /app
RUN git clone --depth 1 --branch master https://github.com/ggml-org/llama.cpp.git
WORKDIR /app/llama.cpp
RUN cmake -B build \
    -DCMAKE_BUILD_TYPE=Release \
    -DGGML_OPENBLAS=ON \
    -DGGML_NATIVE=ON \
    -DGGML_AVX2=ON \
    -DGGML_F16C=ON \
    -DGGML_LTO=ON \
    -DLLAMA_BUILD_EXAMPLES=OFF \
    -DLLAMA_BUILD_SERVER=ON \
    -DLLAMA_BUILD_TESTS=OFF \
    -DLLAMA_BUILD_CLI=OFF \
    -DBUILD_SHARED_LIBS=OFF
RUN cmake --build build --config Release -j 2

FROM docker.io/library/ubuntu:22.04 AS runner
ENV DEBIAN_FRONTEND=noninteractive
ENV TZ=UTC
RUN apt-get update && apt-get install -y --no-install-recommends \
    curl ca-certificates libgomp1 libopenblas0 \
    python3 python3-venv \
    && rm -rf /var/lib/apt/lists/*
RUN python3 -m venv /opt/venv && \
    /opt/venv/bin/pip install --no-cache-dir huggingface-hub
ENV PATH="/opt/venv/bin:$PATH"
RUN mkdir -p /data/models
COPY --from=builder /app/llama.cpp/build/bin/llama-server /usr/local/bin/llama-server
COPY config.env /etc/config.env
COPY entrypoint.sh /usr/local/bin/entrypoint.sh
RUN chmod +x /usr/local/bin/entrypoint.sh
EXPOSE 7860
ENTRYPOINT ["/usr/local/bin/entrypoint.sh"]