- torch.cuda.amp.GradScaler(...) => torch.amp.GradScaler("cuda", ...)
- torch.cuda.amp.autocast(...) => torch.amp.autocast("cuda", ...)
Introduction
This is an audio tagging recipe for Audioset. It aims at predicting the sound events of an audio clip.
./RESULTS.md contains the latest results.
Zipformer
| Encoder | Feature type |
|---|---|
| Zipformer | Frame level fbank |