Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Kernel Restart Due to Missing libcudnn_ops_infer.so.8 on Google Colab #1236

Open
shkstar opened this issue Feb 3, 2025 · 3 comments
Open

Comments

@shkstar
Copy link

shkstar commented Feb 3, 2025

I am running faster-whisper on Google Colab with CTranslate2==4.4.0 and CUDA-enabled PyTorch, but I am encountering kernel restarts due to a missing cuDNN library (libcudnn_ops_infer.so.8). Despite Colab having pre-installed CUDA and cuDNN, it seems this specific library is unavailable.

When I attempt to run my transcription workflow, I get the following error, followed by an automatic kernel restart:

Could not load library libcudnn_ops_infer.so.8. Error: libcudnn_ops_infer.so.8: cannot open shared object file: No such file or directory

After this warning, the kernel restarts with:

WARNING:root:kernel 4d7edf90-1ad4-411f-80d8-1f5bbde0ac17 restarted
INFO KernelRestarter: restarting kernel (1/5), keep random ports

It worked if I add [!apt-get install libcudnn8 libcudnn8-dev]
What’s the problem?

@Purfview
Copy link
Contributor

Purfview commented Feb 3, 2025

Last week Colab switched from cudnn v8 to cudnn v9.
Try CTranslate2==4.5.0 or install cudnn v8.

@shkstar
Copy link
Author

shkstar commented Feb 4, 2025

Thank you, it works.

@iamashwin99
Copy link

I confirm that pip install CTranslate2==4.5.0 works on colab.

Side note: for those using whisperX, whisperX==3.3.1 conflicts with CTranslate2==4.5.0, current work around is to first install whisperX and then the CTranslate2 in a separate command

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants