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
This recipe contains various different ASR models trained with Aishell2.
In AISHELL-2, 1000 hours of clean read-speech data from iOS is published, which is free for academic usage. On top of AISHELL-2 corpus, an improved recipe is developed and released, containing key components for industrial applications, such as Chinese word segmentation, flexible vocabulary expension and phone set transformation etc. Pipelines support various state-of-the-art techniques, such as time-delayed neural networks and Lattic-Free MMI objective funciton. In addition, we also release dev and test data from other channels (Android and Mic).
(From AISHELL-2: Transforming Mandarin ASR Research Into Industrial Scale)
./RESULTS.md contains the latest results.
Transducers
There are various folders containing the name transducer in this folder.
The following table lists the differences among them.
| Encoder | Decoder | Comment | |
|---|---|---|---|
pruned_transducer_stateless5 |
Conformer(modified) | Embedding + Conv1d | same as pruned_transducer_stateless5 in librispeech recipe |
The decoder in transducer_stateless is modified from the paper
Rnn-Transducer with Stateless Prediction Network.
We place an additional Conv1d layer right after the input embedding layer.