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

Passing --diarize result always in (speaker ?) #2056

Open
MichelBahl opened this issue Apr 15, 2024 · 1 comment
Open

Passing --diarize result always in (speaker ?) #2056

MichelBahl opened this issue Apr 15, 2024 · 1 comment

Comments

@MichelBahl
Copy link

No matter what stereo mp3 I convert with ffmpeg to wave --diarize did not work.

Does anyone have a working sample mp3/wave file?

I used following command:
./main -di -m models/ggml-medium.bin -f ./myfile.wav

The results look always like:

[00:00:00.000 --> 00:00:03.500] (speaker ?) Say hello to Picnic, the supermarket that comes to your house.
[00:00:03.500 --> 00:00:07.100] (speaker ?) Huge selection, super fresh and always delivered for free.
[00:00:07.100 --> 00:00:09.300] (speaker ?) Try our family recipes as well.
[00:00:09.300 --> 00:00:10.800] (speaker ?) One click and everything is there.
[00:00:10.800 --> 00:00:12.700] (speaker ?) Download the Picnic app now.
[00:00:12.700 --> 00:00:15.000] (speaker ?) The number one for families in Berlin.
[00:00:15.000 --> 00:00:22.700] (speaker ?) 6 Minute English from bbclearningenglish.com
[00:00:22.700 --> 00:00:26.200] (speaker ?) Hello. This is 6 Minute English from BBC Learning English.
[00:00:26.200 --> 00:00:27.000] (speaker ?) I'm Phil.
[00:00:27.000 --> 00:00:28.300] (speaker ?) And I'm Georgie.
[00:00:28.300 --> 00:00:31.800] (speaker ?) We all know how important exercise is to stay fit
[00:00:31.800 --> 00:00:34.200] (speaker ?) and reduce the risk of heart disease.
[00:00:34.200 --> 00:00:36.200] (speaker ?) Do you exercise much, Phil?
[00:00:36.200 --> 00:00:39.300] (speaker ?) I try to. I ride my bike at the weekend.
[00:00:39.300 --> 00:00:43.300] (speaker ?) But to be honest, I do spend a lot of time sitting down.

@ggerganov
Copy link
Owner

--diarize is very basic and works only with stereo audio where each speaker is in a separate channel

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

2 participants