More Languages , How and When ?

[fire17]

Hi there 👋
Wondering how or when it will be possible to get more languages available into the model. For me Hebrew is the common language and would like to process or output voice in my language .

Thanks a lot and have a good one! :)

[erogol]

We don't have an immediate plan to train a Hebrew model.

If you know a public dataset, you can point to it here and someone from the 🐸 community might like to train a model.

(I move this to the discussions as it is out of the Feature Request definition. )

[thewh1teagle]

I would like to add hebrew support too.
Is there some tutorial or showcase of someone who added new language support for that?
I don't mind to create myself the entire dataset for that and do whatever needed for that.

[fire17]

You guys say ~1100 languages...
Where's the list? Couldn't find it
Thanks and all the best!

[fire17]

@erogol ? 🙏
Also any updates on hebrew tts ? Thx and all the best! :)

[erogol]

https://tts.readthedocs.io/en/latest/inference.html#example-text-to-speech-using-fairseq-models-in-1100-languages

[isolveit-aps]

There is a very good quality "BLARK" for Faroese available under BY-CC 4.0. Can someone tell me if I can easily start experimenting with training a Faroese TTS model in Faroese language? Or how would this normally get to reality in Coqui?

[Lazarus-Leonardo]

You do not speak Hebrew, you speak Yiddish. Nobody speaks Hebrew anymore.