Convolutional computing failed on Centos 7

nlp
1

Hello, all.

I meet a strange problem. The codes below work successfully on Win10 ( python 3.7 and PyTorch 1.6). However, when running the same code on Centos 7 (python 3.7 and PyTorch 1.5), I got an error.

>>> import torch
>>> import torch.nn as nn
>>> cnn = nn.Conv2d(1, 100, (3, 300))
>>> input = torch.randn(64, 1, 16, 300)
>>> output = cnn(input)

The detail error message as follow:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/run/anaconda3/envs/DL/lib/python3.7/site-packages/torch/nn/modules/module.py", line 550, in __call__
    result = self.forward(*input, **kwargs)
  File "/run/anaconda3/envs/DL/lib/python3.7/site-packages/torch/nn/modules/conv.py", line 349, in forward
    return self._conv_forward(input, self.weight)
  File "/run/anaconda3/envs/DL/lib/python3.7/site-packages/torch/nn/modules/conv.py", line 346, in _conv_forward
    self.padding, self.dilation, self.groups)
RuntimeError: NNPACK SpatialConvolution_updateOutput failed

Waiting for your kind help!