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

vadOptions default value #925

Open
ngcheeyuan opened this issue Jul 22, 2024 · 3 comments
Open

vadOptions default value #925

ngcheeyuan opened this issue Jul 22, 2024 · 3 comments

Comments

@ngcheeyuan
Copy link

ngcheeyuan commented Jul 22, 2024

Just putting it out here. It's a good idea to tune your VadOptions if you are facing some issues with it.

Working with non-english inputs,

The default value for VadOptions is a little too aggressive, and I personally recommend a longer padding and lower threshold values.

For padding, what I found was that it needed a longer padding because it tends to cut off the first few syllabus of the words if it was too aggressive.

@ooobo
Copy link

ooobo commented Jul 23, 2024

Be nice to hear what values have worked well for other users (for all languages) - maybe a discussion would be better?

@ckgithub2019
Copy link

Just putting it out here. It's a good idea to tune your VadOptions if you are facing some issues with it.

Working with non-english inputs,

The default value for VadOptions is a little too aggressive, and I personally recommend a longer padding and lower threshold values.

For padding, what I found was that it needed a longer padding because it tends to cut off the first few syllabus of the words if it was too aggressive.

a longer padding means a longer "speech_pad_ms"? By default, the speech_pad_ms is 400, should be changed to 800 or even higher? what's your suggestion?

@ngcheeyuan
Copy link
Author

longer padding means a longer "speech_pad_ms"? By default, the speech_pad_ms is 400, should be changed to 800 or even higher?

Yes. I went with 1000ms and 0.4 threshold iirc. Do listen to the audio before and after silerovad is applied.

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