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

Create main_data_processor #1

Closed
Veldhoen opened this issue Feb 1, 2024 · 2 comments · Fixed by #18
Closed

Create main_data_processor #1

Veldhoen opened this issue Feb 1, 2024 · 2 comments · Fixed by #18

Comments

@Veldhoen
Copy link
Member

Veldhoen commented Feb 1, 2024

A module that can be called from the worker.py to do the actual processing

@greenw0lf
Copy link
Collaborator

Related to what I mentioned today, about having support for batching: SYSTRAN/faster-whisper#59

It's still an open issue and implementing it is a bit trickier (and supposedly the performance without batching of faster-whisper is still faster than WhisperX with batching)

@greenw0lf greenw0lf mentioned this issue Apr 9, 2024
@greenw0lf
Copy link
Collaborator

Video I used for testing: https://www.openbeelden.nl/media/628918/G%C3%BClers_Travels_4

It can be downloaded, run through dane-audio-extraction-worker, then the resulting audio can be used for testing

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

Successfully merging a pull request may close this issue.

2 participants