search

Marksdo / Whisper

Batch Local Transcribe Audio/Movie To Text With Whisper AI Model. Keep Privacy Safe!
https://whisper.marksdo.com
17 stars 3 forks source link

The Medium and Large model of whisper (just below this thread, I have downloaded) #31

Open wewesmil opened 5 months ago

wewesmil commented 5 months ago

English audio to Korean (not using API, using the whisper Language: Korean) I have tried with the Medium and Large Whisper model from below this thread.

having issue of keep getting half of the transcribe in english.. please see the attachment.

ScreenShot 2024-04-19 14 38 36 ScreenShot 2024-04-19 14 36 36
wewesmil commented 5 months ago

the above scrennshot was using the Medium and below screenshot is by using Large from below thread that you posted download link.

ScreenShot 2024-04-19 14 41 59

Seems like the whisper transcribe is not working in korean at all from some point..

Marksdo commented 5 months ago

1.use audio's language model (in your tips is should be an English Model) to transcribe

2.then use translate engine to translate english to korean

Marksdo commented 5 months ago

you can use automation step to do translate after transcribe finished.

Deepl has free api token, you can found it in the web, or some free docker image can free use it to translate