error when run demo.py

[TTTJJJWWW]

Traceback (most recent call last): File "demo.py", line 271, in test() File "demo.py", line 178, in test encoder = encoder.cuda() File "/home/iie/.conda/envs/s2v/lib/python2.7/site-packages/torch/nn/modules/module.py", line 258, in cuda return self._apply(lambda t: t.cuda(device)) File "/home/iie/.conda/envs/s2v/lib/python2.7/site-packages/torch/nn/modules/module.py", line 185, in _apply module._apply(fn) File "/home/iie/.conda/envs/s2v/lib/python2.7/site-packages/torch/nn/modules/rnn.py", line 112, in _apply self.flatten_parameters() File "/home/iie/.conda/envs/s2v/lib/python2.7/site-packages/torch/nn/modules/rnn.py", line 105, in flatten_parameters self.batch_first, bool(self.bidirectional)) RuntimeError: CuDNN error: CUDNN_STATUS_SUCCESS

[kduy]

it's your CUDA problem. what's your environment ?

[TTTJJJWWW]

Ubuntu16.04 cudnn=7.1.3 cuda=9.0 pytorch=0.4.1 torchvision=0.2.1 python=2.7

[lelechen63]

It should be your cuda problem. Please check if your cudnn version, cuda and pytorch version.

[TTTJJJWWW]

Maybe, but i have no idea to fix this. I think the version is ok, how to change that? "Ubuntu16.04 cudnn=7.1.3 cuda=9.0 pytorch=0.4.1 torchvision=0.2.1 python=2.7"

[IQ17]

Try to add this line in models.py

torch.backends.cudnn.enabled=False

[Mora-max]

we are facing same problem when we run cuda 11 with python 3 would you please help?

[aseyedia]

Try to add this line in models.py

torch.backends.cudnn.enabled=False

This worked, but now I'm running into this issue:

=======================================
Start to generate images
Traceback (most recent call last):
  File "demo.py", line 272, in <module>
    test()
  File "demo.py", line 235, in test
    fake_lmark = encoder(example_landmark, input_mfcc)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/arta/ATVGnet/code/models.py", line 54, in forward
    example_landmark_f = self.lmark_encoder(example_landmark)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/container.py", line 91, in forward
    input = module(input)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/linear.py", line 55, in forward
    return F.linear(input, self.weight, self.bias)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/functional.py", line 1024, in linear
    return torch.addmm(bias, input, weight.t())
RuntimeError: cublas runtime error : the GPU program failed to execute at /opt/conda/conda-bld/pytorch_1535488076166/work/aten/src/THC/THCBlas.cu:249

Many of the the others online who report having this problem are using Nvidia 2080Ti with CUDA less than 10, whereas I'm using a 3080 with CUDA 11.2.

I am running Ubuntu 20.04, CUDNN 8, CUDA 11.2, python 2.7.

[DanBigioi]

Try to add this line in models.py torch.backends.cudnn.enabled=False

This worked, but now I'm running into this issue:

=======================================
Start to generate images
Traceback (most recent call last):
  File "demo.py", line 272, in <module>
    test()
  File "demo.py", line 235, in test
    fake_lmark = encoder(example_landmark, input_mfcc)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/arta/ATVGnet/code/models.py", line 54, in forward
    example_landmark_f = self.lmark_encoder(example_landmark)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/container.py", line 91, in forward
    input = module(input)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/module.py", line 477, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/modules/linear.py", line 55, in forward
    return F.linear(input, self.weight, self.bias)
  File "/home/arta/anaconda3/envs/py2/lib/python2.7/site-packages/torch/nn/functional.py", line 1024, in linear
    return torch.addmm(bias, input, weight.t())
RuntimeError: cublas runtime error : the GPU program failed to execute at /opt/conda/conda-bld/pytorch_1535488076166/work/aten/src/THC/THCBlas.cu:249

Many of the the others online who report having this problem are using Nvidia 2080Ti with CUDA less than 10, whereas I'm using a 2080 with CUDA 11.2.

I am running Ubuntu 20.04, CUDNN 8, CUDA 11.2, python 2.7.

Did you ever solve this? Im getting the same error on windows 10, CUDA 11.2, python 3.7

[aseyedia]

@DanBigioi I don't remember. I did have a typo though; I am using a 3080 and not a 2080. I think I settled on not having enough memory.