OSError(WinError 126)模块未找到求助：PyTorch shm.dll加载失败

问题重现 #

运行以下YOLO训练代码：

from ultralytics import YOLO

model = YOLO("yolov8n.yaml")  # build a new model from scratch
model.train(data="data.yaml", epochs=1)  # train the model

触发报错：

Traceback (most recent call last):
  File "d:\Python\Projects\main.py", line 1, in <module>
    from ultralytics import YOLO
  File "C:\Users\Asus\AppData\Roaming\Python\Python311\site-packages\ultralytics\__init__.py", line 5, in <module>
    from ultralytics.data.explorer.explorer import Explorer
  File "C:\Users\Asus\AppData\Roaming\Python\Python311\site-packages\ultralytics\data\__init__.py", line 3, in <module>
    from .base import BaseDataset
  File "C:\Users\Asus\AppData\Roaming\Python\Python311\site-packages\ultralytics\data\base.py", line 15, in <module>
    from torch.utils.data import Dataset
  File "C:\Users\Asus\AppData\Roaming\Python\Python311\site-packages\torch\__init__.py", line 141, in <module>
    raise err
OSError: [WinError 126] The specified module could not be found. Error loading "C:\Users\Asus\AppData\Roaming\Python\Python311\site-packages\torch\lib\shm.dll" or one of its dependencies.

解决方案 #

• 安装/更新VC++ Redistributable运行库
  shm.dll依赖微软VC运行库，下载安装对应x64版本的VC Redistributable，完成后重启电脑再尝试运行代码。

• 重新安装PyTorch
  可能是PyTorch安装不完整导致文件缺失，先卸载现有包：

  pip uninstall torch torchvision torchaudio -y
  

  再根据环境重新安装：

  • CPU版本：

    pip install torch torchvision torchaudio
    

  • 带NVIDIA CUDA支持（以CUDA 11.8为例）：

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

• 验证Python与PyTorch版本兼容性
  确保Python 3.11搭配的是支持该版本的PyTorch包，避免安装过旧的PyTorch版本。

• 修复系统文件
  以管理员权限打开命令提示符，运行以下命令修复系统缺失文件：

  sfc /scannow
  

内容的提问来源于stack exchange，提问作者edi99