安装#

本文档介绍了如何手动安装 vllm-ascend。

要求#

  • 操作系统:Linux

  • Python:>= 3.10, < 3.13

  • 配备昇腾 NPU 的硬件设备,通常为 Atlas 800 A2 系列。

  • 软件:

    软件

    支持的版本

    说明

    Ascend HDK

    参考文档 CANN 9.0.0

    CANN 运行必需

    CANN

    == 9.0.0

    vllm-ascend 和 torch-npu 运行必需

    torch-npu

    == 2.10.0

    vllm-ascend 必需,无需手动安装,后续步骤会自动安装。

    torch

    == 2.10.0

    torch-npu 和 vllm 必需,无需手动安装,后续步骤会自动安装。

    NNAL

    == 9.0.0

    libatb.so 必需,用于启用高级张量算子加速

有两种安装方式:

  • 使用 pip:首先手动或通过 CANN 镜像准备环境,然后使用 pip 安装 vllm-ascend

  • 使用 Docker:直接使用 vllm-ascend 预构建的 Docker 镜像。

配置昇腾 CANN 环境#

在安装之前,请确保固件/驱动和 CANN 已正确安装,更多详情请参考 昇腾环境搭建指南

配置硬件环境#

要验证 Ascend NPU 固件和驱动程序是否正确安装,请运行:

npu-smi info

更多详情请参考昇腾环境搭建指南

配置软件环境#

准备软件环境最简单的方法是直接使用 CANN 官方镜像:

备注

CANN 预构建镜像已包含 NNAL(昇腾神经网络加速库),它为高级张量操作提供 libatb.so 支持。使用预构建镜像时无需额外安装。

# Update DEVICE according to your device (/dev/davinci[0-7])
export DEVICE=/dev/davinci7
# Update the vllm-ascend image
export IMAGE=quay.io/ascend/cann:9.0.0-910b-ubuntu22.04-py3.11
docker run --rm \
    --name vllm-ascend-env \
    --shm-size=1g \
    --device $DEVICE \
    --device /dev/davinci_manager \
    --device /dev/devmm_svm \
    --device /dev/hisi_hdc \
    -v /usr/local/dcmi:/usr/local/dcmi \
    -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
    -v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
    -v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
    -v /etc/ascend_install.info:/etc/ascend_install.info \
    -v /root/.cache:/root/.cache \
    -it $IMAGE bash
点击此处查看“手动安装 CANN”

您也可以选择手动安装 CANN:

警告

如果在运行时遇到 “libatb.so not found” 错误,请确保已按照下文的手动安装步骤正确安装了 NNAL。

# Create a virtual environment.
python -m venv vllm-ascend-env
source vllm-ascend-env/bin/activate

# Install required Python packages.
python -m pip install --upgrade pip
pip3 install attrs numpy decorator sympy cffi pyyaml pathlib2 psutil protobuf scipy requests absl-py wheel typing_extensions

# Download and install the CANN package.
wget --header="Referer: https://www.hiascend.com/" https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/CANN/CANN%209.0.0/Ascend-cann-toolkit_9.0.0_linux-"$(uname -i)".run
chmod +x ./Ascend-cann-toolkit_9.0.0_linux-"$(uname -i)".run
./Ascend-cann-toolkit_9.0.0_linux-"$(uname -i)".run --full
source /usr/local/Ascend/ascend-toolkit/set_env.sh

wget --header="Referer: https://www.hiascend.com/" https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/CANN/CANN%209.0.0/Ascend-cann-910b-ops_9.0.0_linux-"$(uname -i)".run
chmod +x ./Ascend-cann-910b-ops_9.0.0_linux-"$(uname -i)".run
./Ascend-cann-910b-ops_9.0.0_linux-"$(uname -i)".run --install

wget --header="Referer: https://www.hiascend.com/" https://ascend-repo.obs.cn-east-2.myhuaweicloud.com/CANN/CANN%209.0.0/Ascend-cann-nnal_9.0.0_linux-"$(uname -i)".run
chmod +x ./Ascend-cann-nnal_9.0.0_linux-"$(uname -i)".run
./Ascend-cann-nnal_9.0.0_linux-"$(uname -i)".run --install

source /usr/local/Ascend/nnal/atb/set_env.sh

如果您使用的是 vllm-ascend 预构建的 Docker 镜像,则无需额外步骤。

完成上述步骤后,即可开始安装 vllmvllm-ascend

使用 Python 安装#

首先安装系统依赖并配置 pip 镜像源:

# Using apt-get with mirror
sed -i 's|ports.ubuntu.com|mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list
apt-get update -y && apt-get install -y gcc g++ cmake libnuma-dev wget git curl jq
# Or using yum
# yum update -y && yum install -y gcc g++ cmake numactl-devel wget git curl jq
# Config pip mirror,only versions 0.11.0 and earlier are supported, if using a version later than 0.11.0, do not execute this command
pip config set global.index-url https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple

[可选] 如果您在 x86 架构机器上操作或使用 torch-npu 开发版本,请配置 pip 的额外索引 (extra-index):

# For torch-npu dev version or x86 machine
pip config set global.extra-index-url "https://download.pytorch.org/whl/cpu/"

接着,您可以从预构建的 wheel 包安装 vllmvllm-ascend

# Install vllm-project/vllm. The newest supported version is v0.20.2.
pip install vllm==0.20.2

# Install vllm-project/vllm-ascend.
pip install \
--extra-index-url https://mirrors.huaweicloud.com/ascend/repos/pypi/variant https://mirrors.huaweicloud.com/ascend/repos/pypi  \
vllm-ascend==0.20.2rc1

uv-wheelnext 安装方式仅下载基于 vllm 的增量部分,从而减少下载量。首先安装 uv-wheelnext 以支持增量 wheel 包:

# install uv-wheelnext
curl -LsSf https://astral.sh/uv/install.sh | sed 's/verify_checksum "$_file"/true/' | INSTALLER_DOWNLOAD_URL=https://wheelnext.astral.sh sh
source $HOME/.local/bin/env
# Install vllm-project/vllm. The newest supported version is v0.20.2.
pip install vllm==0.20.2

# Install vllm-project/vllm-ascend from wheelnext index.
uv pip install --system \
--extra-index-url https://mirrors.huaweicloud.com/ascend/repos/pypi/variant https://mirrors.huaweicloud.com/ascend/repos/pypi   \
vllm-ascend==0.20.2rc1
点击此处查看“从源代码构建”

或者从源代码构建:

备注

要安装 triton-ascend,请运行:

pip install triton-ascend==3.2.1 --extra-index-url https://mirrors.huaweicloud.com/ascend/repos/pypi

如果您通过 uv 安装,请确保在所有其他包安装完成之后再安装 triton-ascend,以避免依赖解析冲突。

# Install vLLM.
git clone --depth 1 --branch v0.20.2 https://github.com/vllm-project/vllm
cd vllm
VLLM_TARGET_DEVICE=empty pip install -e .
cd ..

# Install vLLM Ascend.
git clone --depth 1 --branch v0.20.2rc1 https://github.com/vllm-project/vllm-ascend.git
cd vllm-ascend
git submodule update --init --recursive
pip install -e .
cd ..

如果您正在为 Atlas A3 构建自定义算子,请手动执行 git submodule update --init --recursive,或确保您的环境可以访问互联网。

备注

构建自定义算子需要 gcc/g++ 版本高于 8 且支持 C++17 或更高标准。如果您在使用 pip install -e . 时遇到 torch-npu 版本冲突,请使用 pip install --no-build-isolation -e . 以在系统环境下构建。如果在编译过程中遇到其他问题,通常是因为使用了非预期的编译器,您可以在编译前导出 CXX_COMPILERC_COMPILER 环境变量来指定 g++ 和 gcc 的具体路径。

如果您在仅 CPU 的环境(无法使用 npu-smi)中构建,需要在执行 pip install -e . 之前设置 SOC_VERSION,以便构建过程针对正确的芯片。您可以参考 Dockerfile* 中的默认值,例如:

  • Atlas A2:export SOC_VERSION=ascend910b1

  • Atlas A3:export SOC_VERSION=ascend910_9391

  • Atlas 300I:export SOC_VERSION=ascend310p1

  • Atlas A5:export SOC_VERSION=<以 "ascend950" 开头的值>

使用 Docker 安装#

vllm-ascend 提供了用于部署的 Docker 镜像。您可以直接从镜像仓库 ascend/vllm-ascend 拉取预构建镜像并运行。

支持的镜像如下:

镜像名称

硬件

操作系统

vllm-ascend:v0.20.2rc1

Atlas A2

Ubuntu

vllm-ascend:v0.20.2rc1-openeuler

Atlas A2

openEuler

vllm-ascend:v0.20.2rc1-a3

Atlas A3

Ubuntu

vllm-ascend:v0.20.2rc1-a3-openeuler

Atlas A3

openEuler

vllm-ascend:v0.20.2rc1-310p

Atlas 300I

Ubuntu

vllm-ascend:v0.20.2rc1-310p-openeuler

Atlas 300I

openEuler

点击此处查看“从 Dockerfile 构建”

或从源代码构建镜像:

git clone https://github.com/vllm-project/vllm-ascend.git
cd vllm-ascend
docker build -t vllm-ascend-dev-image:latest -f ./Dockerfile .
# Update --device according to your device (Atlas A2: /dev/davinci[0-7] Atlas A3:/dev/davinci[0-15]).
# Update the vllm-ascend image according to your environment.
# Note you should download the weight to /root/.cache in advance.
export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1
docker run --rm \
    --name vllm-ascend-env \
    --shm-size=1g \
    --net=host \
    --device /dev/davinci0 \
    --device /dev/davinci1 \
    --device /dev/davinci2 \
    --device /dev/davinci3 \
    --device /dev/davinci4 \
    --device /dev/davinci5 \
    --device /dev/davinci6 \
    --device /dev/davinci7 \
    --device /dev/davinci_manager \
    --device /dev/devmm_svm \
    --device /dev/hisi_hdc \
    -v /usr/local/dcmi:/usr/local/dcmi \
    -v /usr/local/Ascend/driver/tools/hccn_tool:/usr/local/Ascend/driver/tools/hccn_tool \
    -v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
    -v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
    -v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
    -v /etc/ascend_install.info:/etc/ascend_install.info \
    -v /root/.cache:/root/.cache \
    -it $IMAGE bash

默认工作目录为 /workspace,vLLM 和 vLLM Ascend 代码位于 /vllm-workspace,并以开发模式 (pip install -e) 安装,以便开发者无需重新安装即可立即应用更改。

额外信息#

验证安装#

创建并运行一个简单的推理测试。example.py 内容示例如下:

from vllm import LLM, SamplingParams

prompts = [
    "Hello, my name is",
    "The president of the United States is",
    "The capital of France is",
    "The future of AI is",
]

# Create a sampling params object.
sampling_params = SamplingParams(temperature=0.8, top_p=0.95)
# Create an LLM.
llm = LLM(model="Qwen/Qwen3-0.6B")

# Generate texts from the prompts.
outputs = llm.generate(prompts, sampling_params)
for output in outputs:
    prompt = output.prompt
    generated_text = output.outputs[0].text
    print(f"Prompt: {prompt!r}, Generated text: {generated_text!r}")

然后运行:

python example.py

如果遇到 Hugging Face 连接错误(例如:We couldn't connect to 'https://huggingface.co' to load the files, and couldn't find them in the cached files.),请运行以下命令以使用 ModelScope 作为替代方案:

export VLLM_USE_MODELSCOPE=True
pip install modelscope
python example.py

此部分显示 vllm 中成功检测到 ascend 平台:

INFO 05-27 11:40:38 [__init__.py:44] Available plugins for group vllm.platform_plugins:
INFO 05-27 11:40:38 [__init__.py:46] - ascend -> vllm_ascend:register
INFO 05-27 11:40:38 [__init__.py:49] All plugins in this group will be loaded. Set `VLLM_PLUGINS` to control which plugins to load.
INFO 05-27 11:40:38 [__init__.py:238] Platform plugin ascend is activated

此部分显示最终输出:

Prompt: 'Hello, my name is', Generated text: ' Lucy and I am an 8 year old who loves to draw and write stories'
Prompt: 'The president of the United States is', Generated text: " a key leader in the federal government, and the president's role in the executive"
Prompt: 'The capital of France is', Generated text: ' a city. What is the capital of France? The capital of France is Paris'
Prompt: 'The future of AI is', Generated text: ' a topic that is being discussed in various contexts. In the business world, AI'

此部分显示离线推理后进程退出,不影响实际推理:

(EngineCore pid=970) INFO 05-12 11:36:00 [core.py:1201] Shutdown initiated (timeout=0)
(EngineCore pid=970) INFO 05-12 11:36:00 [core.py:1224] Shutdown complete
ERROR 05-12 11:36:01 [core_client.py:704] Engine core proc EngineCore died unexpectedly, shutting down client.
sys:1: DeprecationWarning: builtin type swigvarlink has no __module__ attribute

多节点部署#

验证多节点通信#

首先检查物理层连通性,然后验证每个节点,最后验证节点间连通性。

物理层要求#

  • 物理机必须位于同一局域网内,且网络互通。

  • 所有 NPU 均通过光模块连接,且连接状态必须正常。

单节点验证#

在每个节点上依次执行以下命令。结果必须全部为 success 且状态为 UP

 # Check the remote switch ports
 for i in {0..7}; do hccn_tool -i $i -lldp -g | grep Ifname; done 
 # Get the link status of the Ethernet ports (UP or DOWN)
 for i in {0..7}; do hccn_tool -i $i -link -g ; done
 # Check the network health status
 for i in {0..7}; do hccn_tool -i $i -net_health -g ; done
 # View the network detected IP configuration
 for i in {0..7}; do hccn_tool -i $i -netdetect -g ; done
 # View gateway configuration
 for i in {0..7}; do hccn_tool -i $i -gateway -g ; done
 # View NPU network configuration
 cat /etc/hccn.conf
 # Check the remote switch ports
 for i in {0..15}; do hccn_tool -i $i -lldp -g | grep Ifname; done 
 # Get the link status of the Ethernet ports (UP or DOWN)
 for i in {0..15}; do hccn_tool -i $i -link -g ; done
 # Check the network health status
 for i in {0..15}; do hccn_tool -i $i -net_health -g ; done
 # View the network detected IP configuration
 for i in {0..15}; do hccn_tool -i $i -netdetect -g ; done
 # View gateway configuration
 for i in {0..15}; do hccn_tool -i $i -gateway -g ; done
 # View NPU network configuration
 cat /etc/hccn.conf

互联验证#

1.获取 NPU IP 地址#
for i in {0..7}; do hccn_tool -i $i -ip -g | grep ipaddr; done
for i in {0..15}; do hccn_tool -i $i -ip -g | grep ipaddr; done
2.跨节点 PING 测试#
# Execute on the target node (replace with actual IP)
hccn_tool -i 0 -ping -g address x.x.x.x

在每个节点运行容器#

使用 vLLM-ascend 官方容器运行多节点环境效率更高。

在每个节点运行以下命令以启动容器(您应提前将权重下载到 /root/.cache):

# Update the vllm-ascend image
# openEuler:
# export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1-openeuler
# Ubuntu:
# export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1
export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1

# Run the container using the defined variables
# Note if you are running bridge network with docker, Please expose available ports
# for multiple nodes communication in advance
docker run --rm \
--name vllm-ascend \
--net=host \
--shm-size=1g \
--device /dev/davinci0 \
--device /dev/davinci1 \
--device /dev/davinci2 \
--device /dev/davinci3 \
--device /dev/davinci4 \
--device /dev/davinci5 \
--device /dev/davinci6 \
--device /dev/davinci7 \
--device /dev/davinci_manager \
--device /dev/devmm_svm \
--device /dev/hisi_hdc \
-v /usr/local/dcmi:/usr/local/dcmi \
-v /usr/local/Ascend/driver/tools/hccn_tool:/usr/local/Ascend/driver/tools/hccn_tool \
-v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
-v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
-v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
-v /etc/ascend_install.info:/etc/ascend_install.info \
-v /root/.cache:/root/.cache \
-it $IMAGE bash
# Update the vllm-ascend image
# openEuler:
# export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1-a3-openeuler
# Ubuntu:
# export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1-a3
export IMAGE=quay.io/ascend/vllm-ascend:v0.20.2rc1-a3

# Run the container using the defined variables
# Note if you are running bridge network with docker, Please expose available ports
# for multiple nodes communication in advance
docker run --rm \
--name vllm-ascend \
--net=host \
--shm-size=1g \
--device /dev/davinci0 \
--device /dev/davinci1 \
--device /dev/davinci2 \
--device /dev/davinci3 \
--device /dev/davinci4 \
--device /dev/davinci5 \
--device /dev/davinci6 \
--device /dev/davinci7 \
--device /dev/davinci8 \
--device /dev/davinci9 \
--device /dev/davinci10 \
--device /dev/davinci11 \
--device /dev/davinci12 \
--device /dev/davinci13 \
--device /dev/davinci14 \
--device /dev/davinci15 \
--device /dev/davinci_manager \
--device /dev/devmm_svm \
--device /dev/hisi_hdc \
-v /usr/local/dcmi:/usr/local/dcmi \
-v /usr/local/Ascend/driver/tools/hccn_tool:/usr/local/Ascend/driver/tools/hccn_tool \
-v /usr/local/bin/npu-smi:/usr/local/bin/npu-smi \
-v /usr/local/Ascend/driver/lib64/:/usr/local/Ascend/driver/lib64/ \
-v /usr/local/Ascend/driver/version.info:/usr/local/Ascend/driver/version.info \
-v /etc/ascend_install.info:/etc/ascend_install.info \
-v /root/.cache:/root/.cache \
-it $IMAGE bash