RuntimeError: CUDNN_STATUS_EXECUTION_FAILED

[smallworld-network-wupeng]

root@74a4066c1b48:/app/demoCode# python testBatchModel.py /shared/img_list.txt /shared/3d_output/

Prepare image /shared/cute_boy.jpg:
Number of faces detected: 1
CNN Model loaded to regress 3D Shape and Texture!
Loaded the Basel Face Model to write the 3D output!
Processing image: /shared/3d_output/imgs/cute_boy.png
Traceback (most recent call last):
File "testBatchModel.py", line 168, in
features = net(im).data.cpu().numpy()
File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/module.py", line 357, in call
result = self.forward(*input, **kwargs)
File "../CNN/shape_model.py", line 239, in forward
conv1 = self.conv1(conv1_pad)
File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/module.py", line 357, in call
result = self.forward(*input, **kwargs)
File "/usr/local/lib/python2.7/dist-packages/torch/nn/modules/conv.py", line 282, in forward
self.padding, self.dilation, self.groups)
File "/usr/local/lib/python2.7/dist-packages/torch/nn/functional.py", line 90, in conv2d
return f(input, weight, bias)
RuntimeError: CUDNN_STATUS_EXECUTION_FAILED

I have two 2080 card and ubuntu 16.04 but these is a error CUDNN_STATUS_EXECUTION_FAILED when I run program in Dock. I need help

[smallworld-network-wupeng]

It is looks torch 0.3.1 unsupported the 2070 card. But when I update torch to 1.0 the model unsupported torch 1.0 .

[spiketg]

having the same error with the same card in Ubuntu 18.04

[digbose92]

@smallworld-network-wupeng and @spiketg any updates on this issue ?

[anhttran]

Dear,
One simple solution is to disable cudnn in the main code:
torch.backends.cudnn.enabled = False
Hope it help.
Best regards,
Anh

[Frank-Dz]

Same problem.

[Frank-Dz]

With docker image installed, I met this error.

ENVs:

• CUDA 10.2
• RTX 2080
• Ubuntu16
• Driver Version: 440.33.01

[Frank-Dz]

It is looks torch 0.3.1 unsupported the 2070 card. But when I update torch to 1.0 the model unsupported torch 1.0 .

Did you solve this?