记录：ubuntu18.04 cuda11.7 +torch1.13.1安装apex时报错解决办法

最新推荐文章于 2024-06-04 20:58:59 发布

不明所以~

最新推荐文章于 2024-06-04 20:58:59 发布

阅读量620

点赞数 5

分类专栏： Python 文章标签： pytorch

本文链接：https://blog.csdn.net/qq_43758934/article/details/135411036

版权

Python 专栏收录该内容

5 篇文章 0 订阅

订阅专栏

文章描述了在使用NVIDIA的Apex库进行PyTorch项目中的混合精度和分布式训练时遇到的`ModuleNotFoundError`，特别提到缺少`packaging`模块。提供了解决方案链接，指向GitHubissue中的讨论。

摘要由CSDN通过智能技术生成

git clone https://github.com/NVIDIA/apex
cd apex
# if pip >= 23.1 (ref: https://pip.pypa.io/en/stable/news/#v23-1) which supports multiple `--config-settings` with the same key... 
pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --config-settings "--build-option=--cpp_ext" --config-settings "--build-option=--cuda_ext" ./
# otherwise
pip install -v --disable-pip-version-check --no-cache-dir --no-build-isolation --global-option="--cpp_ext" --global-option="--cuda_ext" ./

NVIDIA/apex: A PyTorch Extension: Tools for easy mixed precision and distributed training in Pytorch (github.com)按照官方安装步骤报错如下：

ModuleNotFoundError: No module named 'packaging'
  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: /home/huiwei/miniconda3/envs/earthformer/bin/python3 /home/huiwei/miniconda3/envs/earthformer/lib/python3.9/site-packages/pip/_vendor/pyproject_hooks/_in_process/_in_process.py get_requires_for_build_wheel /tmp/tmpuel1iqe5
  cwd: /tmp/pip-req-build-7az4lcr7
  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.

解决办法：

https://github.com/NVIDIA/apex/issues/1737#issuecomment-1762662648