search

savbell / whisper-writer

💬📝 A small dictation app using OpenAI's Whisper speech recognition model.
GNU General Public License v3.0
244 stars 40 forks source link

Could not load library cudnn_ops_infer64_8.dll. (Windows) #29

Closed KernAlan closed 5 months ago

KernAlan commented 5 months ago 
Could not load library cudnn_ops_infer64_8.dll. Error code 126
Please make sure cudnn_ops_infer64_8.dll is in your library path!

Tried on Windows 10 and 11.

KernAlan commented 5 months ago

Ah, the problem is that device in local_model_options was set to auto by default. I set it to cpu instead and this solved the issue.

If anyone knows where the CUDA dependencies are being called, I will gladly put in some error handling if needed.

savbell commented 5 months ago

Hi, thanks for the bug report! :)

The local model is created here: https://github.com/savbell/whisper-writer/blob/main/src/transcription.py#L27-L31

I imagine the error comes from it auto-defaulting to GPU and not having these libraries installed: https://github.com/SYSTRAN/faster-whisper#requirements

If you wanted to add some error handling, that would be fantastic! Otherwise I'll add a note in the Readme in my next commit. Thanks again!

KernAlan commented 5 months ago

Thank you! Submitted a PR: https://github.com/savbell/whisper-writer/pull/30

This project is wonderful by the way. I plan to use it to make my prompting faster.

savbell commented 5 months ago

Thanks for your help! I've just merged the PR :)

I'm glad you're finding the project useful! It's always cool when something you make just for yourself ends up being something other people find helpful, too!