安装mistral_inference时出现No module named 'torch'错误求助

未使用Conda环境，已通过PyTorch官网提供的适配CUDA 12.4的命令安装PyTorch：

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124

尝试安装依赖Torch的mistral_inference时，出现如下错误（核心报错为ModuleNotFoundError: No module named 'torch'，发生在xformers构建wheel环节）：

PS C:\Users\admin> pip install mistral_inference
Collecting mistral_inference
  Using cached mistral_inference-1.4.0-py3-none-any.whl.metadata (14 kB)
Collecting fire>=0.6.0 (from mistral_inference)
  Using cached fire-0.7.0.tar.gz (87 kB)
  Installing build dependencies ... done
  Getting requirements to build wheel ... done
  Preparing metadata (pyproject.toml) ... done
Collecting mistral_common>=1.4.0 (from mistral_inference)
  Using cached mistral_common-1.4.4-py3-none-any.whl.metadata (4.6 kB)
Collecting pillow>=10.3.0 (from mistral_inference)
  Using cached pillow-10.4.0-cp312-cp312-win_amd64.whl.metadata (9.3 kB)
Requirement already satisfied: safetensors>=0.4.0 in c:\users\admin\appdata\local\programs\python\python312\lib\site-packages (from mistral_inference) (0.4.4)
Collecting simple-parsing>=0.1.5 (from mistral_inference)
  Using cached simple_parsing-0.1.6-py3-none-any.whl.metadata (7.3 kB)
Collecting xformers>=0.0.24 (from mistral_inference)
  Using cached xformers-0.0.28.post1.tar.gz (7.8 MB)
  Installing build dependencies ... done
  Getting requirements to build wheel ... error
  error: subprocess-exited-with-error

  × Getting requirements to build wheel did not run successfully.
  │ exit code: 1
  ╰─> [20 lines of output]
      Traceback (most recent call last):
        File "C:\Users\admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 353, in <module>
          main()
        File "C:\Users\admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 335, in main
          json_out['return_val'] = hook(**hook_input['kwargs'])
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "C:\Users\admin\AppData\Local\Programs\Python\Python312\Lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 118, in get_requires_for_build_wheel
          return hook(config_settings)
                 ^^^^^^^^^^^^^^^^^^^^^
        File "C:\Users\admin\AppData\Local\Temp\pip-build-env-zjzruady\overlay\Lib\site-packages\setuptools\build_meta.py", line 332, in get_requires_for_build_wheel
          return self._get_build_requires(config_settings, requirements=[])
                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        File "C:\Users\admin\AppData\Local\Temp\pip-build-env-zjzruady\overlay\Lib\site-packages\setuptools\build_meta.py", line 302, in _get_build_requires
          self.run_setup()
        File "C:\Users\admin\AppData\Local\Temp\pip-build-env-zjzruady\overlay\Lib\site-packages\setuptools\build_meta.py", line 503, in run_setup
          super().run_setup(setup_script=setup_script)
        File "C:\Users\admin\AppData\Local\Temp\pip-build-env-zjzruady\overlay\Lib\site-packages\setuptools\build_meta.py", line 318, in run_setup
          exec(code, locals())
        File "<string>", line 24, in <module>
      ModuleNotFoundError: No module named 'torch'
      [end of output]

  note: This error originates from a subprocess, and is likely not a problem with pip.
error: subprocess-exited-with-error

× Getting requirements to build wheel did not run successfully.
│ exit code: 1
╰─> See above for output.

note: This error originates from a subprocess, and is likely not a problem with pip.

一个月前遇到过类似问题，当时通过重新安装PyTorch并替换缺失文件解决，但现在重新执行PyTorch安装命令显示依赖已满足，无法解决当前问题。

解决步骤 #

• 先验证当前环境的torch是否正常
  打开终端执行以下命令，确认torch能正常导入且CUDA可用：

  python -c "import torch; print(torch.__version__); print(torch.cuda.is_available())"
  

• 方法1：禁用pip的构建隔离
  直接在安装命令中添加--no-build-isolation参数，让xformers在当前环境中构建，继承已安装的torch：

  pip install mistral_inference --no-build-isolation
  

• 方法2：手动安装预编译的xformers wheel

  1. 根据你的Python版本（比如3.12）和CUDA版本（12.4），找到对应Windows平台的xformers预编译wheel文件
  2. 下载后执行本地安装：

     pip install 下载路径/xformers-0.0.28.post1-cp312-cp312-win_amd64.whl
     

  3. 再执行pip install mistral_inference

• 方法3：强制重新安装torch并清理缓存
  如果torch存在隐性安装问题，执行以下命令彻底重装：

  pip uninstall -y torch torchvision torchaudio
  pip cache purge
  pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu124
  

  重装完成后再次验证torch，然后安装mistral_inference。

内容的提问来源于stack exchange，提问作者419