Dockerfile

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

# Set working directory
WORKDIR /app

# Copy requirements.txt
COPY requirements.txt .

# Install dependencies
RUN pip install --no-cache-dir -r requirements.txt

# Copy application code
COPY . .

# Set environment variables for Hugging Face cache
ENV HF_HOME=/app/cache
ENV TRANSFORMERS_CACHE=/app/cache
# Set to 1 if using quantization and GPU is available
ENV USE_QUANTIZATION=0

# Expose port for FastAPI
EXPOSE 8000

# Run the FastAPI app with uvicorn (updated to use app.py)
CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8000"]