Spaces:
Runtime error
Runtime error
Creating new speaker voices with speaker embedding datasets
#1
by
bethecloud
- opened
Hi all โ hope everyone is having a great weekend. I have been playing around with SpeechT5 today. I am looking to fine-tune the model on a specific speaker embedding to create unique voices. Are there any guides/resources that I should be looking for to accomplish this? Looking to do something similar to Eleven, but open-source and connected to decentralized cloud (storj).
@bethecloud I'm also interested in this training this model on other voices - did you have any luck?