RuntimeError: Expected a ‘cuda‘ device type for generator but found ‘cpu‘

最新推荐文章于 2024-12-31 12:47:32 发布

AI浩

最新推荐文章于 2024-12-31 12:47:32 发布

阅读量2.2k

点赞数 8

分类专栏：疑难问题文章标签： pytorch 深度学习 python

本文为博主原创文章，未经博主允许不得转载。

本文链接：https://blog.csdn.net/hhhhhhhhhhwwwwwwwwww/article/details/124649651

版权

疑难问题专栏收录该内容

99 篇文章

订阅专栏

在尝试使用PyTorch的数据加载器时，遇到了一个RuntimeError，错误信息表明期望的生成器设备类型为'cuda'，但实际上找到的是'cpu'。为了解决这个问题，修改了数据加载器的初始化参数，添加了`generator=torch.Generator(device='cuda')`，确保在多进程数据加载时使用GPU生成器，从而正确地在GPU上运行计算。

摘要生成于 C知道，由 DeepSeek-R1 满血版支持，前往体验 >

    batch_iterator = iter(data_loader)
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\dataloader.py", line 368, in __iter__
    return self._get_iterator()
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\dataloader.py", line 314, in _get_iterator
    return _MultiProcessingDataLoaderIter(self)
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\dataloader.py", line 965, in __init__
    self._reset(loader, first_iter=True)
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\dataloader.py", line 996, in _reset
    self._try_put_index()
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\dataloader.py", line 1230, in _try_put_index
    index = self._next_index()
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\dataloader.py", line 521, in _next_index
    return next(self._sampler_iter)  # may raise StopIteration
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\sampler.py", line 226, in __iter__
    for idx in self.sampler:
  File "D:\ProgramData\Anaconda3\envs\mytorch\lib\site-packages\torch\utils\data\sampler.py", line 122, in __iter__
    yield from torch.randperm(n, generator=generator).tolist()
RuntimeError: Expected a 'cuda' device type for generator but found 'cpu'

将

    data_loader = data.DataLoader(dataset, args.batch_size,
                                  num_workers=args.num_workers,
                                  shuffle=True, collate_fn=detection_collate,
                                  pin_memory=True)

改为

    data_loader = data.DataLoader(dataset, args.batch_size,
                                  num_workers=args.num_workers,
                                  shuffle=True, collate_fn=detection_collate,
                                  pin_memory=True,generator=torch.Generator(device = 'cuda'))