SkybuckFlying
commented
1 month ago I have same problem:
(base) G:\Environments\Anaconda>dir
Volume in drive G is APPLICATION HARDDISK
Volume Serial Number is 0AE8-C66D
Directory of G:\Environments\Anaconda
24/10/2024 22:48 <DIR> .
24/10/2024 22:34 <DIR> ..
24/10/2024 22:36 0 .conda_envs_dir_test
24/10/2024 22:45 <DIR> apex
24/10/2024 22:41 <DIR> nemo
1 File(s) 0 bytes
4 Dir(s) 10,257,359,077,376 bytes free
(base) G:\Environments\Anaconda>conda activate nemo
(G:\Environments\Anaconda\nemo) G:\Environments\Anaconda>cd apex
(G:\Environments\Anaconda\nemo) G:\Environments\Anaconda\apex>pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" .
Using pip 24.2 from G:\Environments\Anaconda\nemo\lib\site-packages\pip (python 3.10)
DEPRECATION: --build-option and --global-option are deprecated. pip 25.0 will enforce this behaviour change. A possible replacement is to use --config-settings. Discussion can be found at https://github.com/pypa/pip/issues/11859
WARNING: Implying --no-binary=:all: due to the presence of --build-option / --global-option.
Processing g:\environments\anaconda\apex
Running command pip subprocess to install build dependencies
Using pip 24.2 from G:\Environments\Anaconda\nemo\Lib\site-packages\pip (python 3.10)
Collecting setuptools
Using cached setuptools-75.2.0-py3-none-any.whl
Collecting wheel
Using cached wheel-0.44.0-py3-none-any.whl
Installing collected packages: wheel, setuptools
Creating G:\Temp\pip-build-env-g9c7o8b5\overlay\Scripts
Successfully installed setuptools-75.2.0 wheel-0.44.0
Installing build dependencies ... done
Running command Getting requirements to build wheel
Traceback (most recent call last):
File "G:\Environments\Anaconda\nemo\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py", line 353, in <module>
main()
File "G:\Environments\Anaconda\nemo\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 "G:\Environments\Anaconda\nemo\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 "G:\Temp\pip-build-env-g9c7o8b5\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 "G:\Temp\pip-build-env-g9c7o8b5\overlay\Lib\site-packages\setuptools\build_meta.py", line 302, in _get_build_requires
self.run_setup()
File "G:\Temp\pip-build-env-g9c7o8b5\overlay\Lib\site-packages\setuptools\build_meta.py", line 318, in run_setup
exec(code, locals())
File "<string>", line 10, in <module>
ModuleNotFoundError: No module named 'torch'
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.
full command: 'G:\Environments\Anaconda\nemo\python.exe' 'G:\Environments\Anaconda\nemo\lib\site-packages\pip\_vendor\pyproject_hooks\_in_process\_in_process.py' get_requires_for_build_wheel 'G:\Temp\tmpise3elvp'
cwd: G:\Environments\Anaconda\apex
Getting requirements to build wheel ... error
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.
(G:\Environments\Anaconda\nemo) G:\Environments\Anaconda\apex>cd..
(G:\Environments\Anaconda\nemo) G:\Environments\Anaconda>pip install numpy
Requirement already satisfied: numpy in g:\environments\anaconda\nemo\lib\site-packages (1.23.5)
(G:\Environments\Anaconda\nemo) G:\Environments\Anaconda>pip install setuptools
Requirement already satisfied: setuptools in g:\environments\anaconda\nemo\lib\site-packages (75.1.0)
(G:\Environments\Anaconda\nemo) G:\Environments\Anaconda>cd apex
Bing co-pilot AI recommends following possible solution, not tried yet:
If torch still isn't recognized, let's add it manually to the build environment:
Modify Setup Script:
Open the setup.py file in the apex directory and ensure it includes torch in its requirements.
Example of Adding PyTorch to setup.py
python
Copy
import setuptools
from torch.utils.cpp_extension import BuildExtension, CUDAExtension
setuptools.setup(
name='apex',
ext_modules=[
CUDAExtension('apex', [
'apex.cpp',
]),
],
cmdclass={
'build_ext': BuildExtension
},
install_requires=[
'torch>=1.0', # Ensure the version matches your installed version
]
)
Finally, retry the installation:
sh
Copy
pip install -v --no-cache-dir --global-option="--cpp_ext" --global-option="--cuda_ext" .
These steps should help get everything in order. Let's see if this fixes the issue! 🚀 How's it looking now?
coder109
AlongWY
zslefour
尝试了很多此，成功安装。环境为：win10，python3.111，torch2.4.1，cuda12.4 使用CMD powershell会失败，不清楚原因。 将储存库clone到本地，然后运行cmd，进入仓库目录
执行 git checkout apex_no_distributed
执行 pip install -v --no-cache-dir ./
终于成功安装