Dockerfile

FROM nvidia/cuda:11.0.3-base-ubuntu20.04

ENV MODEL_NAME="llava-1.6-mistral-7b-gguf"
ENV DEFAULT_MODEL_FILE="llava-v1.6-mistral-7b.Q3_K_XS.gguf"
ENV MODEL_USER="cjpais"
ENV DEFAULT_MODEL_BRANCH="main"
ENV DEFAULT_CLIP_MODEL_FILE="mmproj-model-f16.gguf"
ENV MODEL_URL="https://huggingface.co/${MODEL_USER}/${MODEL_NAME}/resolve/${DEFAULT_MODEL_BRANCH}/${DEFAULT_MODEL_FILE}"
ENV CLIP_MODEL_URL="https://huggingface.co/${MODEL_USER}/${MODEL_NAME}/resolve/${DEFAULT_MODEL_BRANCH}/${DEFAULT_CLIP_MODEL_FILE}"
ENV PATH="/usr/local/cuda/bin:$PATH"

RUN apt update && \
    apt install --no-install-recommends -y build-essential python3 python3-pip wget curl git && \
    apt clean && rm -rf /var/lib/apt/lists/*

WORKDIR /app

RUN apt-get install -y wget && \
    wget -qO- "https://cmake.org/files/v3.18/cmake-3.18.0-Linux-x86_64.tar.gz" | tar --strip-components=1 -xz -C /usr/local

COPY requirements.txt ./

RUN pip3 install --upgrade pip && \
    pip3 install -r requirements.txt

# Cloning the llama-cpp-python repository
#RUN git clone https://github.com/abetlen/llama-cpp-python.git
#WORKDIR llama-cpp-python
# Initializing and updating the llama.cpp submodule
#RUN git submodule init && git submodule update

# Building the llama-cpp-python
# RUN FORCE_CMAKE=1 pip install .

# Moving back to /app directory
WORKDIR /app

RUN echo ${MODEL_URL} && \
    wget -O /app/${DEFAULT_MODEL_FILE} ${MODEL_URL}

RUN echo ${CLIP_MODEL_URL} && \
    wget -O /app/${DEFAULT_CLIP_MODEL_FILE} ${CLIP_MODEL_URL}
    
RUN useradd -m -u 1000 user

RUN mkdir -p /home/user/app && \
    mv /app/${DEFAULT_MODEL_FILE} /home/user/app

RUN ls /app/${DEFAULT_MODEL_FILE}

RUN chown -R user:user /home/user/app

USER user
ENV HOME=/home/user \
	PATH=/home/user/.local/bin:$PATH \
    MODEL_NAME=${MODEL_NAME} \
    MODEL_FILE=/home/user/app/${DEFAULT_MODEL_FILE} 

WORKDIR $HOME/app

COPY --chown=user . .

RUN ls -al

EXPOSE 8000

CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "8000"]