Dockerfile

FROM python:3.11-slim

WORKDIR /app

ENV PYTHONUNBUFFERED=1 \
    HF_HOME=/app/cache

# CPU-only torch first (much smaller image than the default CUDA build)
RUN pip install --no-cache-dir torch==2.5.1 --index-url https://download.pytorch.org/whl/cpu

COPY requirements.txt .
RUN pip install --no-cache-dir -r requirements.txt

# Bake the embedding model into the image → fast Space startup, no runtime download
RUN python -c "from sentence_transformers import SentenceTransformer; \
    SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')"

COPY . .

# HF Spaces runs containers as uid 1000
RUN useradd -m -u 1000 user && chown -R user:user /app
USER user

EXPOSE 7860
CMD ["gunicorn", "-b", "0.0.0.0:7860", "--workers", "1", "--threads", "4", "--timeout", "120", "app:app"]