Need help with paralinguistic events
#2
by
Dodome
- opened
<|laughter|> <|sigh|> <|breathing|> <|coughing|> <|throat_clearing|>
These are the five paralinguistic events I found in the SoulX-Podcast technical report.
Are there any other paralinguistic events supported by the model? If yes, where can I find the full list of them?
Also, is there a way to add short pauses (a bit of silence) between lines, especially when the speaker changes? Right now, the next speaker starts immediately after the previous one without any gap.