This medium model trained on Common Voice 16 also works
#1
by
Pendrokar
- opened
Medium model: https://huggingface.co/arturslogins/whisper-medium-lv
Tested on forked space: https://huggingface.co/spaces/Pendrokar/LatvianSpeechRecognition/commits/main
Even as a safetensor.
Glad to see extra models popping up. A strange thing I noticed in my tests with this model is that it seems to work much slower than OpenAI base medium model.