FROM python:3.11-slim AS runtime # Install curl for container health checks and keep the image lean. RUN apt-get update \ && apt-get install -y --no-install-recommends curl \ && rm -rf /var/lib/apt/lists/* RUN useradd --create-home appuser WORKDIR /app COPY requirements.txt . RUN pip install --no-cache-dir -r requirements.txt COPY server.py . # Create a zero-byte placeholder so Docker sees the expected mount path. # At runtime this file is replaced by the volume-mounted blitzkode.gguf. RUN touch /app/blitzkode.gguf \ && chown -R appuser:appuser /app ENV BLITZKODE_HOST=0.0.0.0 \ BLITZKODE_PORT=7860 \ BLITZKODE_MODEL_PATH=/app/blitzkode.gguf \ BLITZKODE_GPU_LAYERS=0 \ BLITZKODE_THREADS=4 \ BLITZKODE_THREADS_BATCH=4 \ BLITZKODE_PRELOAD_MODEL=true \ BLITZKODE_N_CTX=2048 \ BLITZKODE_BATCH=256 \ BLITZKODE_UBATCH=128 \ BLITZKODE_PROMPT_CACHE=true \ BLITZKODE_PROMPT_CACHE_BYTES=67108864 EXPOSE 7860 HEALTHCHECK --interval=30s --timeout=10s --start-period=90s --retries=3 \ CMD curl -sf http://localhost:7860/health \ || python -c "import urllib.request; urllib.request.urlopen('http://localhost:7860/health')" \ || exit 1 USER appuser CMD ["python", "server.py"]