collabora / WhisperLive

A nearly-live implementation of OpenAI's Whisper.
MIT License
2.1k stars 286 forks source link

[question]How to get better accuracy for real time english speak to text #238

Open zeliang3 opened 5 months ago

zeliang3 commented 5 months ago

Assuming that hardware performance limitations such as graphics cards\CPUs are not taken into account

  1. Replacement of stronger models, if yes, could you any recommend
  2. Adjustment parameters. if yes, could you any recommend
  3. so on...

Bro, Could you have any suggestions, many thanks!!!