PyTorch RuntimeError: DataLoader worker (pid(s) 15332) 意外退

素材狗 2022-01-01 19:40:22 文章分类：Python问题点击数：次

文章标签

PyTorch RuntimeError: DataLoader worker (pid(s) 15332) exited unexpectedly(PyTorch RuntimeError: DataLoader worker (pid(s) 15332) 意外退出)

本文介绍了PyTorch RuntimeError: DataLoader worker (pid(s) 15332) 意外退出的处理方法，对大家解决问题具有一定的参考价值，需要的朋友们下面随着小编来一起学习吧！

问题描述

我是 PyTorch 的初学者，我只是在此网页上尝试一些示例.但由于此错误，我似乎无法运行super_resolution"程序:

I am a beginner at PyTorch and I am just trying out some examples on this webpage. But I can't seem to get the 'super_resolution' program running due to this error:

RuntimeError: DataLoader worker (pid(s) 15332) 意外退出

上网查了一下，发现有人建议将num_workers设置为0.但是如果我这样做，程序会告诉我内存不足(无论是 CPU 还是 GPU):

I searched the Internet and found that some people suggest setting num_workers to 0. But if I do that, the program tells me that I am running out of memory (either with CPU or GPU):

RuntimeError: [enforce fail at ..c10coreCPUAllocator.cpp:72] 数据.DefaultCPUAllocator:内存不足:您试图分配 9663676416 字节.购买新内存！

或

运行时错误:CUDA 内存不足.尝试分配 1024.00 MiB(GPU 0；4.00 GiB 总容量；已分配 2.03 GiB；0 字节空闲；PyTorch 总共保留 2.03 GiB)

我该如何解决这个问题?

How do I fix this?

我在 Win10(64 位)和 pytorch 1.4.0 上使用 python 3.8.

I am using python 3.8 on Win10(64bit) and pytorch 1.4.0.

更完整的错误信息(--cuda 表示使用 GPU，--threads x 表示将 x 传递给 num_worker 参数):

More complete error messages (--cuda means using GPU, --threads x means passing x to the num_worker parameter):

带命令行参数 --upscale_factor 1 --cuda

  File "E:Python38libsite-packages	orchutilsdatadataloader.py", line 761, in _try_get_data
    data = self._data_queue.get(timeout=timeout)
  File "E:Python38libmultiprocessingqueues.py", line 108, in get
    raise Empty
_queue.Empty

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "Z:super_resolutionmain.py", line 81, in <module>
    train(epoch)
  File "Z:super_resolutionmain.py", line 48, in train
    for iteration, batch in enumerate(training_data_loader, 1):
  File "E:Python38libsite-packages	orchutilsdatadataloader.py", line 345, in __next__
    data = self._next_data()
  File "E:Python38libsite-packages	orchutilsdatadataloader.py", line 841, in _next_data
    idx, data = self._get_data()
  File "E:Python38libsite-packages	orchutilsdatadataloader.py", line 808, in _get_data
    success, data = self._try_get_data()
  File "E:Python38libsite-packages	orchutilsdatadataloader.py", line 774, in _try_get_data
    raise RuntimeError('DataLoader worker (pid(s) {}) exited unexpectedly'.format(pids_str))
RuntimeError: DataLoader worker (pid(s) 16596, 9376, 12756, 9844) exited unexpectedly

带命令行参数 --upscale_factor 1 --cuda --threads 0

  File "Z:super_resolutionmain.py", line 81, in <module>
    train(epoch)
  File "Z:super_resolutionmain.py", line 52, in train
    loss = criterion(model(input), target)
  File "E:Python38libsite-packages	orch
nmodulesmodule.py", line 532, in __call__
    result = self.forward(*input, **kwargs)
  File "Z:super_resolutionmodel.py", line 21, in forward
    x = self.relu(self.conv2(x))
  File "E:Python38libsite-packages	orch
nmodulesmodule.py", line 532, in __call__
    result = self.forward(*input, **kwargs)
  File "E:Python38libsite-packages	orch
nmodulesconv.py", line 345, in forward
    return self.conv2d_forward(input, self.weight)
  File "E:Python38libsite-packages	orch
nmodulesconv.py", line 341, in conv2d_forward
    return F.conv2d(input, weight, self.bias, self.stride,
RuntimeError: CUDA out of memory. Tried to allocate 1024.00 MiB (GPU 0; 4.00 GiB total capacity; 2.03 GiB already allocated; 954.35 MiB free; 2.03 GiB reserved in total by PyTorch)