text-generation-webui in docker

Dockerfile

FROM nvidia/cuda:11.8.0-devel-ubuntu22.04

# https://developer.nvidia.com/cuda-gpus
# for a rtx 2060: ARG TORCH_CUDA_ARCH_LIST="7.5"
ARG TORCH_CUDA_ARCH_LIST="${TORCH_CUDA_ARCH_LIST:-3.5;5.0;6.0;6.1;7.0;7.5;8.0;8.6+PTX}"

LABEL maintainer="Your Name <your.email@example.com>"
LABEL description="Docker image for GPTQ-for-LLaMa and Text Generation WebUI"

RUN --mount=type=cache,target=/var/cache/apt,sharing=locked,rw apt-get update && \
    apt-get install --no-install-recommends -y python3-dev libportaudio2 libasound-dev git python3 python3-pip make g++ ffmpeg && \
    rm -rf /var/lib/apt/lists/*

RUN --mount=type=cache,target=/root/.cache/pip,rw \
    python3 -m pip install virtualenv
RUN mkdir /app

WORKDIR /app

ARG WEBUI_VERSION
RUN test -n "${WEBUI_VERSION}" && git reset --hard ${WEBUI_VERSION} || echo "Using provided webui source"

# Create virtualenv
RUN virtualenv /app/venv
RUN --mount=type=cache,target=/root/.cache/pip,rw \
    . /app/venv/bin/activate && \
    python3 -m pip install --upgrade pip setuptools wheel && \
    python3 -m pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118 && \
    python3 -m pip freeze | grep == | sed 's/==/>=/' > /app/constraints.txt && \
    python3 -m pip install -c /app/constraints.txt "xformers>=0.0.22+cu118"

# Install main requirements
COPY requirements.txt /app/requirements.txt
RUN --mount=type=cache,target=/root/.cache/pip,rw \
    . /app/venv/bin/activate && \
    python3 -m pip install -c /app/constraints.txt -r requirements.txt

COPY . /app/

RUN cp /app/venv/lib/python3.10/site-packages/bitsandbytes/libbitsandbytes_cuda118.so /app/venv/lib/python3.10/site-packages/bitsandbytes/libbitsandbytes_cpu.so

# Install extension requirements
RUN --mount=type=cache,target=/root/.cache/pip,rw \
    . /app/venv/bin/activate && \
    for ext in /app/extensions/*/requirements.txt; do \
    cd "$(dirname "$ext")"; \
    python3 -m pip install -c /app/constraints.txt -r requirements.txt; \
    done

ENV CLI_ARGS=""

EXPOSE ${CONTAINER_PORT:-7860} ${CONTAINER_API_PORT:-5000} ${CONTAINER_API_STREAM_PORT:-5005}
CMD . /app/venv/bin/activate && python3 server.py ${CLI_ARGS}