XTTS model issues

[schmurtzm]

At first run, in XTTS model, the model "base" doesn't work due to a path error:
FileNotFoundError: [Errno 2] No such file or directory: 'C:/tts-webui-installer/base/config.json'
When you refresh the models list it create a second "base" model in the list which is working. Each time you will have to select the second "base" item in the model list to make XTTS work properly.

XTTS is incredible the result is impressive but is there a way to tune it like chatterbox ? (at least tune the speaking speed).

Installing "XTTS Fine-tuning Demo Extension" will broke XTTS (only saying strange sounds after that). No idea how to repair XTTS after even after removing "XTTS Fine-tuning Demo Extension". I had to re-install fully TTS-WebUI after that so if you have a tips to explain how to reset or reinstall a model it would be useful.

Tested on Windows 10 x64.

Some other questions while I'm here :

• May be Speak To Text deserve a dedicated main tab ? With Whisper (currently in Tools section).
• ... And adding Parakeet v3 to this STT section 🙂
• Bark requires an updated tutorial, the UI is complex, not sure how to use it

Thanks for this extraordinary tool

[rsxdalv]

Thanks for the detailed response!

The truth is XTTS (Tortoise derivative) and Bark are old and usually wouldn't be recommended over vibevoice, firered etc, hence why I barely touch them.

I might fix it but not sure about the timeline yet.