mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-12-09 14:05:33 +00:00
3e4da5f781
* Replace deprecated pytorch methods
- torch.cuda.amp.GradScaler(...) => torch.amp.GradScaler("cuda", ...)
- torch.cuda.amp.autocast(...) => torch.amp.autocast("cuda", ...)
* Replace `with autocast(...)` with `with autocast("cuda", ...)`
Co-authored-by: Li Peng <lipeng@unisound.ai>
Introduction
./RESULTS.md contains the latest results.
Transducers
These are the types of architectures currently available.
| Encoder | Decoder | Comment | |
|---|---|---|---|
pruned_transducer_stateless7_streaming |
Streaming Zipformer | Embedding + Conv1d | Adapted from librispeech pruned_transducer_stateless7_streaming |