BELLE-2

/

Belle-whisper-large-v3-zh

like 94

Follow

BELLE-2 Group // Be Everyone's Large Language model Engine 25

I want to use the model for a domain-specific audio set. Is there a way other than fine-tuning the model to make the model recognize domain-specific terms? Does the model support prompt-based fine-tuning? For example, the original whisper API has an initial_prompt parameter for prompting.