Add WhisperSpeech Medium TTS/Voice Cloning model

Open JamesClarke7283 opened this issue 1 year ago • 1 comments

WhisperSpeech has a Medium model now:

https://huggingface.co/WhisperSpeech/WhisperSpeech

It might be more accurate.

Oct 03 '24 04:10 JamesClarke7283

Also it looks that new models support 7 languages!

Unfortunately, these new models do not currently work with the engine code that was published here. The WhisperSpeech team needs to release an update for their project.

Oct 05 '24 16:10 mkiol

The new version 4.7.0 is out and available on flathub.

The new version includes:

New WhisperSpeech Small model for: English, Italian, German, French, Spanish, Dutch and Portuguese

I didn't enable the "Medium" model because I wasn't able to test it properly. It seems that "Medium" has very high GPU memory requirements, and my not-so-new graphics card was not able to run it :/

Dec 29 '24 14:12 mkiol