RUN git clone https://github.com/ggerganov/llama.cpp RUN pip install gguf -i https://pypi.tuna.tsinghua.edu.cn/simple WORKDIR /llama.cpp RUN mkdir build WORKDIR /llama.cpp/build RUN cmake .. -DLLAMA_CUDA=ON RUN cmake --build . --config Release # python build RUN CMAKE_ARGS="-...