| # Grab a fresh copy of the Python image | |
| FROM python:3.10-slim | |
| # Install build and runtime dependencies | |
| RUN apt-get update && \ | |
| apt-get install -y \ | |
| libopenblas-dev \ | |
| ninja-build \ | |
| build-essential \ | |
| curl | |
| RUN pip install -U pip setuptools wheel && \ | |
| CMAKE_ARGS="-DLLAMA_OPENBLAS=on" FORCE_CMAKE=1 pip install llama-cpp-python[server] | |
| # Download model | |
| RUN mkdir model && \ | |
| curl -L https://huggingface.co/TheBloke/h2ogpt-4096-llama2-13B-chat-GGML/blob/main/h2ogpt-4096-llama2-13b-chat.ggmlv3.q6_K.bin -o model/ggmlv3-model.bin | |
| COPY ./start_server.sh ./start_server.sh | |
| # Make the server start script executable | |
| RUN chmod +x ./start_server.sh | |
| # Set environment variable for the host | |
| ENV HOST=0.0.0.0 | |
| ENV PORT=7860 | |
| # Expose a port for the server | |
| EXPOSE ${PORT} | |
| # Run the server start script | |
| CMD ["/bin/sh", "./start_server.sh"] |