PyTorch says NO cuda GPU found on Colab however torch.cuda.is_available() is True

Asked Jan 29 '21 at 13:04

Active Jan 30 '21 at 09:47

Viewed 1,803 times

Is there any way around to get GPU on colab

import torch
print(torch.version)
print(torch.cuda.is_available())

Error:

RuntimeError: No CUDA GPUs are available

Any suggestion would be appreciated

edited Jan 30 '21 at 09:47

Subbu VidyaSekar

asked Jan 29 '21 at 13:04

1

Does this answer your question? [How to activate GPU computing in Google colab?](https://stackoverflow.com/questions/61210095/how-to-activate-gpu-computing-in-google-colab) – Ivan Jan 29 '21 at 13:12
I already activated GPU by using shell script thats why torch.cuda.is_available() is returning true. – Muhammad Usman Rasheed Jan 29 '21 at 13:17
Ok, when is this error raised? – Ivan Jan 29 '21 at 13:20

0 Answers0