Dockerfile

FROM python:3.11-slim

RUN apt-get update && apt-get install -y --no-install-recommends     cmake build-essential git git-lfs curl     && rm -rf /var/lib/apt/lists/*     && git lfs install

WORKDIR /app

RUN pip install --no-cache-dir --upgrade pip &&     pip install --no-cache-dir         "huggingface_hub==0.23.4"         "gradio==4.44.1"         "uvicorn>=0.14.0"         "fastapi<0.113.0"

# Compile llama-cpp-python from source for glibc
RUN CMAKE_ARGS="-DLLAMA_BLAS=OFF -DLLAMA_CUBLAS=OFF -DLLAMA_METAL=OFF"     pip install --no-cache-dir "llama-cpp-python==0.3.2"

COPY app.py .

ENV PYTHONUNBUFFERED=1
ENV GRADIO_SERVER_NAME=0.0.0.0
ENV GRADIO_SERVER_PORT=7860

EXPOSE 7860
CMD ["python", "app.py"]