Change to whisper-large-v3-turbo model

#10
by newton41 - opened

OpenAI has released the new whisper model turbo, which is about half size of original large model. I knew that the whisper-jax-spaces is running super fast now, but i am wondering adapting the new turbo model can lower the TPU request with just a little bit performance losing. So, may I ask how to change this space model setting from large to turbo? Thank you.

Sign up or log in to comment