Looks like they left out all training code, presumably for commercial reasons (but it only just came out so it's conceivable they are just cleaning up that side of the code but I doubt it). Totally their call, given they've put the effort in, just a shame.
Comments
They have shared the link to GitHub [1], HuggingFace repo [2], and the paper [3]:
1. https://github.com/aiola-lab/whisper-medusa
2. https://huggingface.co/aiola/whisper-medusa-v1
3. https://paperswithcode.com/method/multi-head-attention
Looks like they left out all training code, presumably for commercial reasons (but it only just came out so it's conceivable they are just cleaning up that side of the code but I doubt it). Totally their call, given they've put the effort in, just a shame.