# default base image
# https://gallery.ecr.aws/neuron/pytorch-inference-neuronx
ARG BASE_IMAGE="public.ecr.aws/neuron/pytorch-inference-neuronx:2.5.1-neuronx-py310-sdk2.22.0-ubuntu22.04"

FROM $BASE_IMAGE

RUN echo "Base image is $BASE_IMAGE"

# Install some basic utilities
RUN apt-get update && \
    apt-get install -y \
        git \
        python3 \
        python3-pip \
        ffmpeg libsm6 libxext6 libgl1

### Mount Point ###
# When launching the container, mount the code directory to /workspace
ARG APP_MOUNT=/workspace
VOLUME [ ${APP_MOUNT} ]
WORKDIR ${APP_MOUNT}/vllm

RUN python3 -m pip install --upgrade pip
RUN python3 -m pip install --no-cache-dir fastapi ninja tokenizers pandas tenacity
RUN python3 -m pip install sentencepiece transformers==4.48.0 -U
RUN python3 -m pip install neuronx-cc==2.17.194.0 --extra-index-url=https://pip.repos.neuron.amazonaws.com -U
RUN python3 -m pip install pytest

# uninstall transformers-neuronx package explicitly to avoid version conflict
RUN python3 -m pip uninstall -y transformers-neuronx

COPY . .
ARG GIT_REPO_CHECK=0
RUN --mount=type=bind,source=.git,target=.git \
    if [ "$GIT_REPO_CHECK" != 0 ]; then bash tools/check_repo.sh ; fi

RUN python3 -m pip install -U \
        'cmake>=3.26' ninja packaging 'setuptools-scm>=8' wheel jinja2 \
        -r requirements/neuron.txt

ENV VLLM_TARGET_DEVICE neuron
RUN --mount=type=bind,source=.git,target=.git \
    pip install --no-build-isolation -v -e .

# install development dependencies (for testing)
RUN python3 -m pip install -e tests/vllm_test_utils

# install transformers-neuronx package as an optional dependencies (for V0)
# FIXME: `--no-deps` argument is temporarily added to resolve transformers package version conflict
RUN python3 -m pip install transformers-neuronx==0.13.* --extra-index-url=https://pip.repos.neuron.amazonaws.com -U --no-deps

# overwrite entrypoint to run bash script
RUN echo "import subprocess; import sys; subprocess.check_call(sys.argv[1:])" > /usr/local/bin/dockerd-entrypoint.py

CMD ["/bin/bash"]