What kind of voice models TVM support currently?

Questions
#1

Does anybody know TVM support which public voice task models? Have list ? The easy one is KWS model, but like ASR models , streaming models…etc. They have more complex architectures, like input dymanic shape, cache …

#2

@leshenj15 have some experience in using TVM to accelerate whisper.

1 Like
#3

@yzh119 , Could you share some tutorial of how did you do that? Thanks.