Yes. The example uses Whisper-tiny which is 39M, a perfect match for the downstream task of keyword spotting. Just one line needs to be changed in the code to r

Yes. The example uses Whisper-tiny which is 39M, a perfect match for the downstream task of keyword spotting. Just one line needs to be changed in the code to run a larger Whisper model :)