[BUG] Installation doesn't work (Unable to import torch, pre-compiling ops will be disabled.)

Describe the bug
I can't install deepspeed using "pip install deepspeed". It says that I don't have pytorch, eventhough I have pytorch.

To Reproduce
Steps to reproduce the behavior:

1. Install python 3.10.11
2. Check your python version

D:\IA\DeepSpeed>python --version
Python 3.10.11

3. Create a venv

D:\IA\DeepSpeed>python -m venv venv

4. Activate the venv

D:\IA\DeepSpeed>venv\Scripts\activate

5. Upgrade pip (important, or else it won't install pytorch)

(venv) D:\IA\DeepSpeed>python -m pip install --upgrade pip
Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com
Requirement already satisfied: pip in d:\ia\deepspeed\venv\lib\site-packages (23.0.1)
Collecting pip
  Downloading pip-25.1.1-py3-none-any.whl (1.8 MB)
     ---------------------------------------- 1.8/1.8 MB 38.6 MB/s eta 0:00:00
Installing collected packages: pip
  Attempting uninstall: pip
    Found existing installation: pip 23.0.1
    Uninstalling pip-23.0.1:
      Successfully uninstalled pip-23.0.1
Successfully installed pip-25.1.1

6. Install pytorch 2.7.0 with cuda 12.8

(venv) D:\IA\DeepSpeed>pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu128
Looking in indexes: https://download.pytorch.org/whl/cu128, https://pypi.ngc.nvidia.com
Collecting torch
  Downloading https://download.pytorch.org/whl/cu128/torch-2.7.0%2Bcu128-cp310-cp310-win_amd64.whl.metadata (29 kB)
Collecting torchvision
  Downloading https://download.pytorch.org/whl/cu128/torchvision-0.22.0%2Bcu128-cp310-cp310-win_amd64.whl.metadata (6.3 kB)
Collecting torchaudio
  Downloading https://download.pytorch.org/whl/cu128/torchaudio-2.7.0%2Bcu128-cp310-cp310-win_amd64.whl.metadata (6.8 kB)
Collecting filelock (from torch)
  Downloading https://download.pytorch.org/whl/filelock-3.13.1-py3-none-any.whl.metadata (2.8 kB)
Collecting typing-extensions>=4.10.0 (from torch)
  Downloading https://download.pytorch.org/whl/typing_extensions-4.12.2-py3-none-any.whl.metadata (3.0 kB)
Collecting sympy>=1.13.3 (from torch)
  Downloading https://download.pytorch.org/whl/sympy-1.13.3-py3-none-any.whl.metadata (12 kB)
Collecting networkx (from torch)
  Downloading https://download.pytorch.org/whl/networkx-3.3-py3-none-any.whl.metadata (5.1 kB)
Collecting jinja2 (from torch)
  Downloading https://download.pytorch.org/whl/Jinja2-3.1.4-py3-none-any.whl.metadata (2.6 kB)
Collecting fsspec (from torch)
  Downloading https://download.pytorch.org/whl/fsspec-2024.6.1-py3-none-any.whl.metadata (11 kB)
Collecting numpy (from torchvision)
  Downloading https://download.pytorch.org/whl/numpy-2.1.2-cp310-cp310-win_amd64.whl.metadata (59 kB)
Collecting pillow!=8.3.*,>=5.3.0 (from torchvision)
  Downloading https://download.pytorch.org/whl/pillow-11.0.0-cp310-cp310-win_amd64.whl.metadata (9.3 kB)
Collecting mpmath<1.4,>=1.1.0 (from sympy>=1.13.3->torch)
  Downloading https://download.pytorch.org/whl/mpmath-1.3.0-py3-none-any.whl (536 kB)
     ---------------------------------------- 536.2/536.2 kB ? eta 0:00:00
Collecting MarkupSafe>=2.0 (from jinja2->torch)
  Downloading https://download.pytorch.org/whl/MarkupSafe-2.1.5-cp310-cp310-win_amd64.whl (17 kB)
Downloading https://download.pytorch.org/whl/cu128/torch-2.7.0%2Bcu128-cp310-cp310-win_amd64.whl (3338.3 MB)
   ---------------------------------------- 3.3/3.3 GB 77.6 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/cu128/torchvision-0.22.0%2Bcu128-cp310-cp310-win_amd64.whl (7.6 MB)
   ---------------------------------------- 7.6/7.6 MB 79.3 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/cu128/torchaudio-2.7.0%2Bcu128-cp310-cp310-win_amd64.whl (4.7 MB)
   ---------------------------------------- 4.7/4.7 MB 93.4 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/pillow-11.0.0-cp310-cp310-win_amd64.whl (2.6 MB)
   ---------------------------------------- 2.6/2.6 MB 144.0 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/sympy-1.13.3-py3-none-any.whl (6.2 MB)
   ---------------------------------------- 6.2/6.2 MB 47.4 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/typing_extensions-4.12.2-py3-none-any.whl (37 kB)
Downloading https://download.pytorch.org/whl/filelock-3.13.1-py3-none-any.whl (11 kB)
Downloading https://download.pytorch.org/whl/fsspec-2024.6.1-py3-none-any.whl (177 kB)
Downloading https://download.pytorch.org/whl/Jinja2-3.1.4-py3-none-any.whl (133 kB)
Downloading https://download.pytorch.org/whl/networkx-3.3-py3-none-any.whl (1.7 MB)
   ---------------------------------------- 1.7/1.7 MB 90.0 MB/s eta 0:00:00
Downloading https://download.pytorch.org/whl/numpy-2.1.2-cp310-cp310-win_amd64.whl (12.9 MB)
   ---------------------------------------- 12.9/12.9 MB 53.6 MB/s eta 0:00:00
Installing collected packages: mpmath, typing-extensions, sympy, pillow, numpy, networkx, MarkupSafe, fsspec, filelock, jinja2, torch, torchvision, torchaudio
Successfully installed MarkupSafe-2.1.5 filelock-3.13.1 fsspec-2024.6.1 jinja2-3.1.4 mpmath-1.3.0 networkx-3.3 numpy-2.1.2 pillow-11.0.0 sympy-1.13.3 torch-2.7.0+cu128 torchaudio-2.7.0+cu128 torchvision-0.22.0+cu128 typing-extensions-4.12.2

7. Here is my environnement:

(venv) D:\IA\DeepSpeed>pip list
Package           Version
----------------- ------------
filelock          3.13.1
fsspec            2024.6.1
Jinja2            3.1.4
MarkupSafe        2.1.5
mpmath            1.3.0
networkx          3.3
numpy             2.1.2
pillow            11.0.0
pip               25.1.1
setuptools        65.5.0
sympy             1.13.3
torch             2.7.0+cu128
torchaudio        2.7.0+cu128
torchvision       0.22.0+cu128
typing_extensions 4.12.2

8. Check that pytorch is properly installed

(venv) D:\IA\DeepSpeed>python
Python 3.10.11 (tags/v3.10.11:7d4cc5a, Apr  5 2023, 00:38:17) [MSC v.1929 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import torch
>>> torch.__version__
'2.7.0+cu128'
>>> torch.cuda.is_available()
True
>>> exit()

9. Install deepspeed using pip

(venv) D:\IA\DeepSpeed>pip install deepspeed
Looking in indexes: https://pypi.org/simple, https://pypi.ngc.nvidia.com
Collecting deepspeed
  Downloading deepspeed-0.16.7.tar.gz (1.5 MB)
     ---------------------------------------- 1.5/1.5 MB 40.3 MB/s eta 0:00:00
  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]
      [WARNING] Unable to import torch, pre-compiling ops will be disabled. Please visit https://pytorch.org/ to see how to properly install torch on your system.
      ←[93m [WARNING] ←[0m unable to import torch, please install it if you want to pre-compile any deepspeed ops.
      DS_BUILD_OPS=1
      Traceback (most recent call last):
        File "D:\IA\DeepSpeed\venv\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 389, in <module>
          main()
        File "D:\IA\DeepSpeed\venv\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 373, in main
          json_out["return_val"] = hook(**hook_input["kwargs"])
        File "D:\IA\DeepSpeed\venv\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 143, in get_requires_for_build_wheel
          return hook(config_settings)
        File "C:\Users\Armaguedin\AppData\Local\Temp\pip-build-env-e2l_hvis\overlay\Lib\site-packages\setuptools\build_meta.py", line 331, in get_requires_for_build_wheel
          return self._get_build_requires(config_settings, requirements=[])
        File "C:\Users\Armaguedin\AppData\Local\Temp\pip-build-env-e2l_hvis\overlay\Lib\site-packages\setuptools\build_meta.py", line 301, in _get_build_requires
          self.run_setup()
        File "C:\Users\Armaguedin\AppData\Local\Temp\pip-build-env-e2l_hvis\overlay\Lib\site-packages\setuptools\build_meta.py", line 512, in run_setup
          super().run_setup(setup_script=setup_script)
        File "C:\Users\Armaguedin\AppData\Local\Temp\pip-build-env-e2l_hvis\overlay\Lib\site-packages\setuptools\build_meta.py", line 317, in run_setup
          exec(code, locals())
        File "<string>", line 157, in <module>
      AssertionError: Unable to pre-compile ops without torch installed. Please install torch before attempting to pre-compile ops.
      [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.

Expected behavior
No error

ds_report output
ds_report doesn't work since deepseek installation failed.

Screenshots
If applicable, add screenshots to help explain your problem.

System info

• OS: Windows 10 Pro
• GPU count and types: A single RTX 5090
• No interconnection
• Python 3.10.11
• Driver Version: 572.83
• CUDA Version: 12.8

Additional informations
I tried to install deepspeed manually but it quickly failed (developer command prompt is running in administrator)

**********************************************************************
** Visual Studio 2022 Developer Command Prompt v17.12.2
** Copyright (c) 2022 Microsoft Corporation
**********************************************************************

C:\Windows\System32>D:

D:\>cd ia

D:\IA>cd DeepSpeed

D:\IA\DeepSpeed>venv\Scripts\activate

(venv) D:\IA\DeepSpeed>build_win.bat
D:\IA\DeepSpeed\venv\Scripts\python.exe: No module named build

The report's entry points are pip install deepspeed and build_win.bat; begin by reproducing the Python 3.10.11/Windows virtual-environment installation and compare the successful torch import with the build output. Done means deepspeed installs without the torch assertion, and the Windows build path no longer stops with the reported missing build module.