SpeechT5 tts, CMU Arctic speaker embeddings

[User110112]

Hello,
I can't get the speecht5 tts option to work properly.
No matter what embedding name (bdl, slt,...) I set there is no change. Only a young woman's voice with much grain, noice in the background.
Am I using wrong embedding names ? Is there something I am not paying attention to ?
Have a great day.

[silentoplayz]

Ensure the voice name has been set in two places:

1. Settings -> Audio tab:
   
   2: Admin Panel -> Settings -> Audio tab:
   

I will see about updating the documentation regarding this if this helps you!

[User110112]

@silentoplayz thank your for the answer. For some reason the voice does not change, no matter what speaker name I use. It's always the woman's voice.
Do you have further recommendations ?

[richtong]

Can't reproduce on 0.5.3 on MacOS running pipx installed though. Seems to work fine with OpenAI

[orrinwitt]

I'm having the same issue, and I think @User110112 is referring to the voices available when using the local (transformers) TTS

[ItsJustRuby]

Specifying the full name of the dataset as TTS Model worked for me, e.g. cmu_us_clb_arctic-wav-arctic_a0001.