Files
enginex-bi_series-vllm/Dockerfile
2025-08-05 19:02:46 +08:00

20 lines
957 B
Docker

FROM zibo.harbor.iluvatar.com.cn:30000/saas/bi100-3.2.1-x86-ubuntu20.04-py3.10-poc-llm-infer:20250731115755
RUN pip install --no-cache-dir triton==2.1.0
COPY pkgs/triton /usr/local/corex/lib64/python3/dist-packages/triton
COPY pkgs/triton-2.1.0+corex.4.1.2.dist-info /usr/local/corex/lib64/python3/dist-packages/triton-2.1.0+corex.4.1.2.dist-info
COPY pkgs/xformers-0.0.22+corex.4.1.2.dist-info /usr/local/corex/lib64/python3/dist-packages/xformers-0.0.22+corex.4.1.2.dist-info
COPY pkgs/xformers /usr/local/corex/lib64/python3/dist-packages/xformers
COPY paged_attn.py /usr/local/lib/python3.10/site-packages/vllm/attention/ops/paged_attn.py
COPY __init__.py /usr/local/lib/python3.10/site-packages/vllm/triton_utils/__init__.py
COPY prefix_prefill.py /usr/local/lib/python3.10/site-packages/vllm/attention/ops/prefix_prefill.py
RUN mkdir /workspace
WORKDIR /workspace/
COPY ./launch_service /workspace/launch_service
ENTRYPOINT ["./launch_service"]