Spaces:
Running
on
Zero
Running
on
Zero
SSML Tags
#1
by
Hatman
- opened
I noticed that VibeVoice is patterned after Microsoft's Azure AI Speech and it'll take the same SSML Tags . Would be cool if the space allowed for that.
As per VibeVoice docs it does not support SSML like Azure Speech. It is built to work with plain text and the model picks up roles and expression from context itself
That sounds cool, any chance to get a link? I haven't seen that on the GitHub and guess I don't really know where to look for the docs.