possibility of using something other than cosyvoice?
#5
by
Grimble
- opened
Is the use of cosyvoice baked into the model, or would it be straightforward to use another voice generator?
we only use cosyvoice the generation audio input data for textQA pairs, you can use any other multi-speaker tts system as you like to generate the data.
gpt-omni
changed discussion status to
closed