Dockerfile

# Use a lightweight Python 3.10 base image
FROM python:3.10-slim

# 1. Hugging Face strictly requires running as a non-root user (UID 1000)
RUN useradd -m -u 1000 user
USER user

# 2. Set up environment variables for the user
ENV HOME=/home/user \
    PATH=/home/user/.local/bin:$PATH

# 3. Set the working directory
WORKDIR $HOME/app

# 4. Copy requirements first to leverage Docker layer caching
COPY --chown=user requirements.txt $HOME/app/

# 5. Install dependencies, using the pre-built CPU wheel to bypass compilation
RUN pip install --no-cache-dir --upgrade pip && \
    pip install --no-cache-dir -r requirements.txt \
    --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cpu

# 6. Copy the main application code
COPY --chown=user app.py $HOME/app/

# 7. Expose port 7860 (The standard port for HF Spaces)
EXPOSE 7860

# 8. Start the Uvicorn server
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]