alphacep / vosk-server

WebSocket, gRPC and WebRTC speech recognition server based on Vosk and Kaldi libraries
Apache License 2.0
871 stars 241 forks source link

Reducing vocabulary #217

Closed olevanss closed 1 year ago

olevanss commented 1 year ago

Hi everyone

I am looking for a way to improve transcribation quality. To achieve it I want to test cutting the vocabulary down. I have found about dynamic vocabulary change of small models.

My question: Is there a way to do it for big models? In particular, russian big model? It contains HCLG.fst and is static.

Thank you!

nshmyrev commented 1 year ago

Same as https://github.com/alphacep/vosk-api/issues/1327