Dockerfile

# Use a lightweight Python image optimized for ML workloads
FROM python:3.10-slim

# Set working directory
WORKDIR /app

# Set environment variables for better performance
ENV PYTHONUNBUFFERED=1 \
    PYTHONDONTWRITEBYTECODE=1 \
    DEBIAN_FRONTEND=noninteractive

# Install system-level dependencies
RUN apt-get update && \
    apt-get install -y \
    ffmpeg \
    libsndfile1 \
    wget \
    curl \
    git \
    build-essential \
    && rm -rf /var/lib/apt/lists/*

# Create non-root user to avoid git permission issues
RUN useradd -m appuser && chown -R appuser /app

# Set HOME explicitly to match the non-root user
ENV HOME=/home/appuser

# Switch to appuser for git config
USER appuser

# Configure git to avoid permission issues during HF build
RUN git config --global --add safe.directory /app && \
    git config --global user.email "[email protected]" && \
    git config --global user.name "peace2024"

# Switch back to root to install dependencies
USER root

# Copy requirements first to leverage Docker layer caching
COPY requirements-hf.txt .

# Upgrade pip and install Python dependencies
RUN pip install --no-cache-dir --upgrade pip && \
    pip install --no-cache-dir -r requirements-hf.txt

# Copy the entire app source code
COPY . .

# Create necessary directories
RUN mkdir -p vector_store logs

# Expose port 7860 (used by Hugging Face Spaces)
EXPOSE 7860

# Install PyTorch with CUDA support (will fall back to CPU if no GPU)
RUN pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cpu

# Health check
HEALTHCHECK --interval=30s --timeout=30s --start-period=5s --retries=3 \
    CMD curl -f http://localhost:7860/docs || exit 1

# Run the FastAPI app via Uvicorn
CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "7860", "--workers", "1"]