| # Use Python 3.9 slim image for better performance | |
| FROM python:3.9-slim | |
| # Set working directory | |
| WORKDIR /app | |
| # Set environment variables | |
| ENV PYTHONUNBUFFERED=1 | |
| ENV PYTHONDONTWRITEBYTECODE=1 | |
| ENV TRANSFORMERS_CACHE=/app/cache | |
| ENV HF_HOME=/app/cache | |
| # Install system dependencies | |
| RUN apt-get update && apt-get install -y \ | |
| build-essential \ | |
| curl \ | |
| git \ | |
| && rm -rf /var/lib/apt/lists/* | |
| # Copy requirements first for better caching | |
| COPY requirements.txt . | |
| # Create a non-root user early | |
| RUN useradd --create-home --shell /bin/bash app | |
| # Install Python dependencies as app user | |
| RUN pip install --no-cache-dir --upgrade pip && \ | |
| pip install --no-cache-dir -r requirements.txt | |
| # Create cache directory and set permissions | |
| RUN mkdir -p /app/cache && \ | |
| chown -R app:app /app | |
| # Copy application code and set ownership | |
| COPY --chown=app:app app.py . | |
| # Switch to non-root user | |
| USER app | |
| # Pre-download models at build time (optional - comment out if you want faster builds) | |
| # RUN python -c " | |
| # from transformers import AutoTokenizer, AutoModel; | |
| # models = ['Lyon28/Albert-Base-V2', 'Lyon28/GPT-2', 'Lyon28/Tinny-Llama']; | |
| # [AutoTokenizer.from_pretrained(m) for m in models[:3]]; | |
| # [AutoModel.from_pretrained(m) for m in models[:3]]; | |
| # print('Sample models cached') | |
| # " | |
| # Expose port | |
| EXPOSE 7860 | |
| # Health check | |
| HEALTHCHECK --interval=30s --timeout=30s --start-period=60s --retries=3 \ | |
| CMD curl -f http://localhost:7860/health || exit 1 | |
| # Run the application | |
| CMD ["python", "app.py"] |