TTS Custom Voice

[shiffman]

In #45, I added support for Xenova/speecht5_tts, need to look into how to train custom speaker embeddings.

Starting point: microsoft/SpeechT5#16

Thanks again @xenova!

[ylacombe]

Hey @shiffman, I don't know if you have already found this resource, but it seems like it could help you extract speaker embeddings from a single extract (without training)!