collabora / WhisperLive

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

Delay in transcription #210

Open alpcansoydas opened 6 months ago

alpcansoydas commented 6 months ago

When the speech time is long and words are continuous, delay goes little bit higher. The model runs on Nvidia T4 GPU. Should I consider upgrading my GPU model or any idea about it?

makaveli10 commented 6 months ago

@alpcansoydas we havent tested it on a T4 yet. Consider using a smaller size model and see if the issue persists.