老显卡老cpu用vllm推理大模型失败Intel(R) Xeon(R) CPU E5-2643 v2

skywalk81632025-11-05 11:06

先上结论，显卡太老，无法装cuda12.6

cpu太老，不支持AVX2, ，所以实践失败。

先安装vllm

复制代码 
pip install vllm

它会把torch一起安装

老显卡的驱动cuda是11.6，尝试升级

复制代码 
sudo apt install nvidia-driver-580-server

没升上去。

尝试升级驱动

看看驱动

复制代码 
ubuntu-drivers devices

但是没有看到有啥驱动，显示：ERROR:root:aplay command not found

如果没有这个工具，就可以安装

安装工具

复制代码 
sudo apt install ubuntu-drivers-common

查询可用驱动

复制代码 
ubuntu-drivers devices

没看到有啥

复制代码 
ubuntu-drivers devices
ERROR:root:aplay command not found

安装cuda11.6的torch

复制代码 
pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116

报错没有

pip install torch==1.13.1+cu116 torchvision==0.14.1+cu116 torchaudio==0.13.1 --extra-index-url https://download.pytorch.org/whl/cu116

Looking in indexes: https://pypi.tuna.tsinghua.edu.cn/simple, https://download.pytorch.org/whl/cu116

ERROR: Could not find a version that satisfies the requirement torch==1.13.1+cu116 (from versions: 2.2.0, 2.2.1, 2.2.2, 2.3.0, 2.3.1, 2.4.0, 2.4.1, 2.5.0, 2.5.1, 2.6.0, 2.7.0, 2.7.1, 2.8.0, 2.9.0)

ERROR: No matching distribution found for torch==1.13.1+cu116

安装cuda 11.8的

复制代码 
pip install torch==2.0.0 torchvision==0.15.1 torchaudio==2.0.1 --index-url https://download.pytorch.org/whl/cu118

报错

pip install torch==2.0.0 torchvision==0.15.1 torchaudio==2.0.1 --index-url https://download.pytorch.org/whl/cu118

Looking in indexes: https://download.pytorch.org/whl/cu118

ERROR: Could not find a version that satisfies the requirement torch==2.0.0 (from versions: 2.2.0+cu118, 2.2.1+cu118, 2.2.2+cu118, 2.3.0+cu118, 2.3.1+cu118, 2.4.0+cu118, 2.4.1+cu118, 2.5.0+cu118, 2.5.1+cu118, 2.6.0+cu118, 2.7.0+cu118, 2.7.1+cu118)

ERROR: No matching distribution found for torch==2.0.0

也就是只有这几个版本：2.2.0+cu118, 2.2.1+cu118, 2.2.2+cu118, 2.3.0+cu118, 2.3.1+cu118, 2.4.0+cu118, 2.4.1+cu118, 2.5.0+cu118, 2.5.1+cu118, 2.6.0+cu118, 2.7.0+cu118, 2.7.1+cu118

安装2.7.1试试

复制代码 
pip install torch==2.7.0 torchvision==0.22 torchaudio==2.7.1 --index-url https://down
load.pytorch.org/whl/cu118

经过反复调试，最终命令是：

复制代码 
pip install torch==2.7.0 torchvision==0.22 torchaudio==2.7.0 --index-url https://download.pytorch.org/whl/cu118

还是一样的报错

尝试升级cuda和cudnn

复制代码 
sudo apt upgrade  nvidia-cuda-dev nvidia-cudnn

碰到了这个报错

sudo apt upgrade nvidia-cuda-dev nvidia-cudnn

Reading package lists... Done

Building dependency tree... Done

Reading state information... Done

You might want to run 'apt --fix-broken install' to correct these.

The following packages have unmet dependencies:

intel-oneapi-runtime-dpcpp-cpp : Depends: intel-oneapi-runtime-compilers (>= 2025.3.0-639) but 2025.0.4-1519 is installed

intel-oneapi-runtime-opencl : Depends: intel-oneapi-runtime-compilers (>= 2025.3.0-639) but 2025.0.4-1519 is installed

E: Unmet dependencies. Try 'apt --fix-broken install' with no packages (or specify a solution).

安装提示执行

复制代码 
sudo apt --fix-broken install

重新编译vllm cpu版

下载

复制代码 
git clone https://gitcode.com/GitHub_Trending/vl/vllm vllm_source

构建

复制代码 
pip install --upgrade pip
pip install -v -r requirements/cpu-build.txt --extra-index-url https://download.pytorch.org/whl/cpu
pip install -v -r requirements/cpu.txt --extra-index-url https://download.pytorch.org/whl/cpu

构筑后端

复制代码 
VLLM_TARGET_DEVICE=cpu python setup.py install

报错

复制代码 
running build_ext
-- The CXX compiler identification is GNU 13.3.0
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++ - skipped
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Build type: RelWithDebInfo
-- Target device: cpu
-- Found Python: /home/skywalk/py312/bin/python (found version "3.12.3") found components: Interpreter Development.Module Development.SABIModule
-- Found python matching: /home/skywalk/py312/bin/python.
CMake Warning at /home/skywalk/py312/lib/python3.12/site-packages/torch/share/cmake/Torch/TorchConfig.cmake:22 (message):
  static library kineto_LIBRARY-NOTFOUND not found.
Call Stack (most recent call first):
  /home/skywalk/py312/lib/python3.12/site-packages/torch/share/cmake/Torch/TorchConfig.cmake:125 (append_torchlib_if_found)
  CMakeLists.txt:84 (find_package)


-- Found Torch: /home/skywalk/py312/lib/python3.12/site-packages/torch/lib/libtorch.so
CMake Error at cmake/cpu_extension.cmake:188 (message):
  vLLM CPU backend requires AVX512, AVX2, Power9+ ISA, S390X ISA, ARMv8 or
  RISC-V support.
Call Stack (most recent call first):
  CMakeLists.txt:104 (include)


-- Configuring incomplete, errors occurred!
Traceback (most recent call last):
  File "/home/skywalk/github/vllm_source/setup.py", line 706, in <module>
    setup(
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/__init__.py", line 117, in setup
    return distutils.core.setup(**attrs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/core.py", line 186, in setup
    return run_commands(dist)
           ^^^^^^^^^^^^^^^^^^
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/core.py", line 202, in run_commands
    dist.run_commands()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1002, in run_commands
    self.run_command(cmd)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/dist.py", line 1104, in run_command
    super().run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1021, in run_command
    cmd_obj.run()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/command/install.py", line 109, in run
    self.do_egg_install()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/command/install.py", line 167, in do_egg_install
    self.run_command('bdist_egg')
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/cmd.py", line 357, in run_command
    self.distribution.run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/dist.py", line 1104, in run_command
    super().run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1021, in run_command
    cmd_obj.run()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/command/bdist_egg.py", line 177, in run
    cmd = self.call_command('install_lib', warn_dir=False)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/command/bdist_egg.py", line 163, in call_command
    self.run_command(cmdname)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/cmd.py", line 357, in run_command
    self.distribution.run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/dist.py", line 1104, in run_command
    super().run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1021, in run_command
    cmd_obj.run()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/command/install_lib.py", line 19, in run
    self.build()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/command/install_lib.py", line 113, in build
    self.run_command('build_ext')
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/cmd.py", line 357, in run_command
    self.distribution.run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/dist.py", line 1104, in run_command
    super().run_command(command)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/dist.py", line 1021, in run_command
    cmd_obj.run()
  File "/home/skywalk/github/vllm_source/setup.py", line 282, in run
    super().run()
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/command/build_ext.py", line 99, in run
    _build_ext.run(self)
  File "/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_distutils/command/build_ext.py", line 368, in run
    self.build_extensions()
  File "/home/skywalk/github/vllm_source/setup.py", line 239, in build_extensions
    self.configure(ext)
  File "/home/skywalk/github/vllm_source/setup.py", line 216, in configure
    subprocess.check_call(
  File "/usr/lib/python3.12/subprocess.py", line 413, in check_call
    raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['cmake', '/home/skywalk/github/vllm_source', '-G', 'Ninja', '-DCMAKE_BUILD_TYPE=RelWithDebInfo', '-DVLLM_TARGET_DEVICE=cpu', '-DVLLM_PYTHON_EXECUTABLE=/home/skywalk/py312/bin/python', '-DVLLM_PYTHON_PATH=/home/skywalk/github/vllm_source:/usr/lib/python312.zip:/usr/lib/python3.12:/usr/lib/python3.12/lib-dynload:/home/skywalk/py312/lib/python3.12/site-packages:/home/skywalk/github/exo:/home/skywalk/py312/lib/python3.12/site-packages/setuptools/_vendor', '-DFETCHCONTENT_BASE_DIR=/home/skywalk/github/vllm_source/.deps', '-DCMAKE_JOB_POOL_COMPILE:STRING=compile', '-DCMAKE_JOB_POOLS:STRING=compile=24']' returned non-zero exit status 1.

没办法了，cpu不行，Intel(R) Xeon(R) CPU E5-2643 v2 @ 3.50GHz 太老了。

使用预编译的cpu版本

复制代码 
pip install vllm --no-build-isolation

白搭，这个也不行，放弃。

vllm -V

W1030 18:34:21.393673118 OperatorEntry.cpp:218 Warning: Warning only once for all operators, other operators may also be overridden.

Overriding a previously registered kernel for the same operator and the same dispatch key

operator: aten::_addmm_activation(Tensor self, Tensor mat1, Tensor mat2, *, Scalar beta=1, Scalar alpha=1, bool use_gelu=False) -> Tensor

registered at /pytorch/build/aten/src/ATen/RegisterSchema.cpp:6

dispatch key: AutocastCPU

previous kernel: registered at /pytorch/aten/src/ATen/autocast_mode.cpp:327

new kernel: registered at /opt/workspace/ipex-cpu-dev/csrc/cpu/autocast/autocast_mode.cpp:112 (function operator())

ERROR! Intel® Extension for PyTorch* only works on machines with instruction sets equal or newer than AVX2, which are not detected on the current machine.

这台机器放弃！

相关推荐
sell-cloud9 分钟前
Prompt Engineering要凉了？Fable 5.5能“猜“意图，但银行正在封杀AI Agent——企业AI的真正护城河是什么
人工智能·prompt
Navicat中国2 小时前
Navicat 18 内置 AI 助手让数据库开发从繁琐操作迈向智能协作
大数据·人工智能·数据库开发
4SAPI2 小时前
多模型请求路由怎么设计？成本、速度和能力如何平衡
人工智能·python·gateway·php
魔乐社区2 小时前
翼芯・智算测试与适配优化专区，让国产算力“选得准、跑得通、用得好“
人工智能·开源·大模型
XCDoris4 小时前
郴州logo设计机构
大数据·人工智能·python
Goboy4 小时前
一个人带三个Agent,我用全新TRAE从0到1搭了个资讯周刊站
人工智能·trae
俊哥V4 小时前
每日 AI 研究简报 · 2026-10-19
人工智能·ai
AI_AGENT_DEV_AI4 小时前
AI英语教培智能体的开发
人工智能
物联网软硬件开发-轨物科技5 小时前
【技术白皮书】光伏清检一体化机器人系统解决方案技术白皮书
大数据·运维·网络·人工智能·机器人
gCode Teacher 格码致知5 小时前
Pandas教学-17:Series与Dataframe的关系
开发语言·python·pandas
热门推荐
01国内可直接用、免费额度/永久免费的大模型API清单（含 SiliconFlow、火山、阿里、智谱、百度、Kimi、DeepSeek、DMXAPI 等）02如何新建文件夹？ 电脑新建文件夹的4种方法03历年考研数学一、数学二、数学三真题试卷及答案PDF04GitHub 镜像站点05几个好用的ip纯净度检测网站06「全语录+省流版」邵艾伦 × 孙宇晨：聊风口、聊个人IP、聊年轻人如何抓住 AI 时代的机会07苹果iPA游戏软件资源下载网站08昇腾910B系列芯片完全指南：四款型号对比、服务器选型与大模型部署实战09Windows 10/11 如何退出微软账户并改用本地账户登录（含找不到选项的解决方法）10蜘蛛磁力 搜索引擎大全，如何使用蜘蛛磁力查找磁力链接