commitcopilot/infer-002

1FROM python:3.11-slim2 3ENV PYTHONDONTWRITEBYTECODE=1 \4    PYTHONUNBUFFERED=1 \5    PORT=7860 \6    LLAMA_SERVER_HOST=127.0.0.1 \7    LLAMA_SERVER_PORT=8080 \8    LD_LIBRARY_PATH=/opt/llama.cpp9 10WORKDIR /app11 12RUN apt-get update && \13    apt-get install -y --no-install-recommends \14    ca-certificates \15    curl \16    gzip \17    libgomp1 \18    libstdc++6 \19    tar && \20    rm -rf /var/lib/apt/lists/*21 22COPY scripts/install_llama_cpp.py ./scripts/install_llama_cpp.py23RUN python scripts/install_llama_cpp.py24 25COPY requirements.txt ./26RUN pip install --no-cache-dir --upgrade pip setuptools wheel && \27    pip install --no-cache-dir -r requirements.txt28 29COPY config.json ./config.json30COPY scripts/download_model.py ./scripts/download_model.py31 32ARG HF_TOKEN=""33RUN --mount=type=secret,id=HF_TOKEN,mode=0444,required=false \34    sh -c 'TOKEN="${HF_TOKEN}"; \35    if [ -f /run/secrets/HF_TOKEN ]; then TOKEN="$(cat /run/secrets/HF_TOKEN)"; fi; \36    python scripts/download_model.py \37    --config config.json \38    --output-dir /models \39    --hf-token "${TOKEN}"'40 41COPY app.py ./42 43EXPOSE 786044CMD ["sh", "-c", "uvicorn app:app --host 0.0.0.0 --port ${PORT:-7860}"]45