Open souzatharsis opened 5 days ago
SSML is definitively interesting. Azure speech services supports it too. But in the context of a generative AI project, do we expect the user to edit the transcript to add these markups, or are there way to generate or enrich an existing transcript with these markups? Do anyone has some success stories about leveraging gen AI this way?
Anyway, that was more a sidetrack than a real meaningful comment. The more tts we have, the better :)
The advantage there is more SSML support and cheaper.
Suggested by Gilgamesh from NotebookLM Discord.