TTS model documentation update

[lumpidu]

TTS model documentation update

A clear and concise description of what the bug is.

The Readme's for the currently supported TTS models are outdated. E.g. there is no hint on that now VITS is supported, nor how SSL-Fastpitch is to be used and how it improves/deviates from the normal Fastpitch model.

Please list minimal steps or code snippet for us to be able to reproduce the bug

Navigate to the TTS description: https://docs.nvidia.com/deeplearning/nemo/user-guide/docs/en/main/tts/intro.html. Search for VITS docuementation or SSL Fastpitch.

Expected behavior

Thorough documentation of the new language models or how to train and if possible current status and performance metrics.

[XuesongYang]

could you please have a look? Thanks.
@redoctopus for SSL;
@treacker for VITS.

[redoctopus]

Both of these are still under development as @experimental models, so we don't guarantee that they won't change (and break backwards compatibility) in the short term.

We could add a section in the documentation regarding current experimental models/models under development, though that may also fall behind main occasionally. @XuesongYang thoughts?

[github-actions]

This issue is stale because it has been open for 30 days with no activity. Remove stale label or comment or this will be closed in 7 days.

[XuesongYang]

we added some docs relevant to the above models though they are still experimental. Closing this issue for now, but please feel free to re-open it anytime.