Siggmoid's picture
Update scoring: MS MARCO embeddings and skill-centric semantic matching
1905876
raw
history blame contribute delete
970 Bytes
# Hugging Face Spaces Docker: https://huggingface.co/docs/hub/spaces-sdks-docker
FROM python:3.11-slim
# Required for Hugging Face Spaces Dev Mode
RUN useradd -m -u 1000 user
WORKDIR /app
# System deps for PyMuPDF
RUN apt-get update && apt-get install -y --no-install-recommends \
libglib2.0-0 \
libsm6 \
libxrender1 \
libxext6 \
&& rm -rf /var/lib/apt/lists/*
ENV HF_HOME=/home/user/.cache/huggingface \
PORT=7860
RUN mkdir -p ${HF_HOME} && chown -R user:user /home/user
COPY --chown=user requirements.txt requirements.txt
RUN pip install --no-cache-dir -r requirements.txt
# Pre-download embedding model at build time (faster Space cold start)
RUN su - user -c "python -c \"from sentence_transformers import SentenceTransformer; SentenceTransformer('msmarco-distilbert-base-v4')\""
COPY --chown=user . /app
USER user
ENV PATH=/home/user/.local/bin:$PATH
EXPOSE 7860
CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]