EngineX/xc-llm-kunlun
3
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

[Doc] update base image url（1.Replace conda with uv; 2.Integrate xpytorch and ops into the image.) (#146)

Browse Source 
Signed-off-by: WeiJie_Hong <1462519292@qq.com>
This commit is contained in:
WeiJie_Hong
2026-01-23 18:55:56 +08:00
committed by GitHub
parent c0f06d04b1
commit 2a998286c0
2 changed files with 27 additions and 63 deletions
Download Patch File Download Diff File Expand all files Collapse all files

42
docs/Dockerfile.xpu
View File

@@ -30,33 +30,40 @@ RUN apt-get update && \
gcc g++ cmake libnuma-dev \
wget tmux curl \
software-properties-common && \
apt remove -y python3.8 python3.8-minimal python3.8-dev && \
apt purge -y python3.8 python3.8-minimal python3.8-dev && \
apt-get clean && \
rm -rf /var/lib/apt/lists/*
RUN conda init && conda create --name vllm_kunlun_0.10.1.1 python=3.10.15 -y && \
conda run -n vllm_kunlun_0.10.1.1 && source activate vllm_kunlun_0.10.1.1 && \
conda clean -afy
RUN sed -i '/# >>> conda initialize >>>/,/# <<< conda initialize <<</d' ~/.bashrc && \
rm -rf /root/miniconda3
RUN source activate vllm_kunlun_0.10.1.1 && \
pip install torch==2.5.1+cu118 torchvision==0.20.1+cu118 torchaudio==2.5.1+cu118 --index-url https://download.pytorch.org/whl/cu118 && \
pip cache purge && rm -rf /root/.cache/pip
ENV PATH=/root/.local/bin/:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
RUN source activate vllm_kunlun_0.10.1.1 && \
pip install setuptools==80.9.0 cuda_mock==1.1.1 hyperparameter==0.5.6 black==23.3.0 lark==1.2.2 \
RUN curl -LsSf https://astral.sh/uv/install.sh | sh && \
source $HOME/.local/bin/env
RUN uv venv --python 3.10 /opt/vllm_kunlun
ENV PATH=/opt/vllm_kunlun/bin:${PATH}
RUN source /opt/vllm_kunlun/bin/activate && \
uv pip install torch==2.5.1+cu118 torchvision==0.20.1+cu118 torchaudio==2.5.1+cu118 --index-url https://download.pytorch.org/whl/cu118 && \
rm -rf /root/.cache/pip && uv cache clean
RUN uv pip install setuptools==80.9.0 cuda_mock==1.1.1 hyperparameter==0.5.6 black==23.3.0 lark==1.2.2 \
networkx wheel PyYAML==6.0.2 ipython h5py regex==2024.9.11 colorama==0.4.6 pynvml==11.5.3 \
nvidia-cuda-runtime-cu11 tabulate==0.9.0 openpyxl==3.1.5 pandas prettytable \
nvidia-cuda-runtime-cu11 tabulate==0.9.0 openpyxl==3.1.5 pandas prettytable setproctitle\
pytest==8.1.0 pytest-repeat==0.9.3 pytest-timeout==2.3.1 py==1.11.0 datasets==2.16.0 \
pydantic==2.9.2 psutil==6.1.0 einops==0.8.0 \
pytest-html==4.1.1 py-cpuinfo pytest-timeout==2.3.1 termcolor jsonlines==4.0.0 tiktoken \
pytest-html==4.1.1 pytest-timeout==2.3.1 termcolor jsonlines==4.0.0 tiktoken \
qwen_vl_utils filetype fastapi==0.112.1 fire py-cpuinfo transformers==4.57.1 \
gradio==4.0.0 sse-starlette trl==0.8.6 uvicorn accelerate==0.30.1 --index-url https://pip.baidu-int.com/simple/ && \
pip cache purge && rm -rf /root/.cache/pip
rm -rf /root/.cache/pip && uv cache clean
ENV DEBIAN_FRONTEND=noninteractive \
LANG=en_US.UTF-8 \
LANGUAGE=en_US:en \
TERM=xterm-256color \
PATH=/root/miniconda/envs/vllm_kunlun_0.10.1.1/bin/:$PATH
TERM=xterm-256color
WORKDIR /workspace
RUN wget https://su.bcebos.com/v1/klx-sdk-release-public/xccl/resource/MLNX_OFED_LINUX-24.01-0.3.3.1-ubuntu20.04-x86_64.tgz && tar -xf MLNX_OFED_LINUX-24.01-0.3.3.1-ubuntu20.04-x86_64.tgz
@@ -67,18 +74,13 @@ RUN apt-get install -y -o Dpkg::Options::='--force-confdef' -o Dpkg::Options::='
apt-get clean && \
rm -rf /var/lib/apt/lists/*
WORKDIR /workspace
COPY vllm-kunlun/ /workspace/vllm-kunlun/
RUN cd /workspace/vllm-kunlun && \
bash dockerfile/install.sh && \
rm -rf /workspace/vllm-kunlun/build/ && rm -rf /workspace/vllm-kunlun/vllm_kunlun.egg-info/
# xpu-smi tools
WORKDIR /workspace
COPY xre-Linux-x86_64-5.2.0.0/ /workspace/xre-Linux-x86_64-5.2.0.0/
RUN mv /workspace/xre-Linux-x86_64-5.2.0.0/bin/* /usr/local/bin/ && mv /workspace/xre-Linux-x86_64-5.2.0.0/so/* /lib/x86_64-linux-gnu/ && \
rm -rf /workspace/xre-Linux-x86_64-5.2.0.0/
ENV LD_LIBRARY_PATH=/opt/vllm_kunlun/lib:/opt/vllm_kunlun/lib64:/lib/x86_64-linux-gnu/:/opt/vllm_kunlun/xcudart/lib/
RUN rm -rf \
/root/.cache \
/root/.conda \

46
docs/source/installation.md
View File

@@ -11,7 +11,7 @@ This document describes how to install vllm-kunlun manually.
- vLLM (same version as vllm-kunlun)
## Setup environment using container
We provide a clean, minimal base image for your use`wjie520/vllm_kunlun:v0.0.1`.You can pull it using the `docker pull` command.
We provide a clean, minimal base image for your use`wjie520/vllm_kunlun:base_v0.0.2` and `wjie520/vllm_kunlun:base_mimo_v0.0.2`(Only MIMO_V2 and GPT-OSS).You can pull it using the `docker pull` command.
### Container startup script
:::::{tab-set}
@@ -31,17 +31,15 @@ if [ $XPU_NUM -gt 0 ]; then
done
DOCKER_DEVICE_CONFIG="${DOCKER_DEVICE_CONFIG} --device=/dev/xpuctrl:/dev/xpuctrl"
fi
export build_image="wjie520/vllm_kunlun:v0.0.1"
export build_image="wjie520/vllm_kunlun:base_v0.0.2"
# or export build_image="iregistry.baidu-int.com/xmlir/xmlir_ubuntu_2004_x86_64:v0.32"
docker run -itd ${DOCKER_DEVICE_CONFIG} \
--net=host \
--cap-add=SYS_PTRACE --security-opt seccomp=unconfined \
--tmpfs /dev/shm:rw,nosuid,nodev,exec,size=32g \
--cap-add=SYS_PTRACE \
-v /home/users/vllm-kunlun:/home/vllm-kunlun \
-v /usr/local/bin/xpu-smi:/usr/local/bin/xpu-smi \
--name "$1" \
-w /workspace \
"$build_image" /bin/bash
@@ -51,10 +49,7 @@ docker run -itd ${DOCKER_DEVICE_CONFIG} \
## Install vLLM-kunlun
### Install vLLM 0.11.0
```
conda activate vllm_kunlun_0.10.1.1
# or conda activate python310_torch25_cuda
pip install vllm==0.11.0 --no-build-isolation --no-deps
uv pip install vllm==0.11.0 --no-build-isolation --no-deps
```
### Build and Install
Navigate to the vllm-kunlun directory and build the package:
@@ -63,45 +58,12 @@ git clone https://github.com/baidu/vLLM-Kunlun
cd vLLM-Kunlun
pip install -r requirements.txt
uv pip install -r requirements.txt
python setup.py build
python setup.py install
```
### Replace eval_frame.py
Copy the eval_frame.py patch:
```
cp vllm_kunlun/patches/eval_frame.py /root/miniconda/envs/vllm_kunlun_0.10.1.1/lib/python3.10/site-packages/torch/_dynamo/eval_frame.py
```
## Install the KL3-customized build of PyTorch
```
wget -O xpytorch-cp310-torch251-ubuntu2004-x64.run https://baidu-kunlun-public.su.bcebos.com/v1/baidu-kunlun-share/1130/xpytorch-cp310-torch251-ubuntu2004-x64.run?authorization=bce-auth-v1%2FALTAKypXxBzU7gg4Mk4K4c6OYR%2F2025-12-02T05%3A01%3A27Z%2F-1%2Fhost%2Ff3cf499234f82303891aed2bcb0628918e379a21e841a3fac6bd94afef491ff7
bash xpytorch-cp310-torch251-ubuntu2004-x64.run
```
## Install the KL3-customized build of PyTorch(Only MIMO V2)
```
wget -O xpytorch-cp310-torch251-ubuntu2004-x64.run https://klx-sdk-release-public.su.bcebos.com/kunlun2aiak_output/1231/xpytorch-cp310-torch251-ubuntu2004-x64.run
bash xpytorch-cp310-torch251-ubuntu2004-x64.run
```
## Install custom ops
```
pip install "https://baidu-kunlun-public.su.bcebos.com/v1/baidu-kunlun-share/1130/xtorch_ops-0.1.2209%2B6752ad20-cp310-cp310-linux_x86_64.whl?authorization=bce-auth-v1%2FALTAKypXxBzU7gg4Mk4K4c6OYR%2F2025-12-05T06%3A18%3A00Z%2F-1%2Fhost%2F14936c2b7e7c557c1400e4c467c79f7a9217374a7aa4a046711ac4d948f460cd"
```
## Install custom ops(Only MIMO V2)
```
pip install "https://vllm-ai-models.bj.bcebos.com/v1/vLLM-Kunlun/ops/swa/xtorch_ops-0.1.2109%252B523cb26d-cp310-cp310-linux_x86_64.whl"
```
## Install the KLX3 custom Triton build
```
pip install "https://cce-ai-models.bj.bcebos.com/v1/vllm-kunlun-0.11.0/triton-3.0.0%2Bb2cde523-cp310-cp310-linux_x86_64.whl"
```
## Install the AIAK custom ops library
```
pip install "https://cce-ai-models.bj.bcebos.com/XSpeedGate-whl/release_merge/20251219_152418/xspeedgate_ops-0.0.0-cp310-cp310-linux_x86_64.whl"
```
## Quick Start