mirror of
https://github.com/k2-fsa/icefall.git
synced 2025-08-09 10:02:22 +00:00
c4ee2bc0af
* a copy of stateless4 as base * distillation with hubert * fix typo * example usage * usage * Update egs/librispeech/ASR/pruned_transducer_stateless6/hubert_xlarge.py Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com> * fix comment * add results of 100hours * Update egs/librispeech/ASR/pruned_transducer_stateless6/hubert_xlarge.py Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com> * Update egs/librispeech/ASR/pruned_transducer_stateless6/hubert_xlarge.py Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com> * check fairseq and quantization * a short intro to distillation framework * Update egs/librispeech/ASR/pruned_transducer_stateless6/hubert_xlarge.py Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com> * add intro of statless6 in README * fix type error of dst_manifest_dir * Update egs/librispeech/ASR/pruned_transducer_stateless6/hubert_xlarge.py Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com> * make export.py call stateless6/train.py instead of stateless2/train.py * update results by stateless6 * adjust results format * fix typo Co-authored-by: Fangjun Kuang <csukuangfj@gmail.com>
30 lines
2.4 KiB
Markdown
30 lines
2.4 KiB
Markdown
# Introduction
|
|
|
|
Please refer to <https://icefall.readthedocs.io/en/latest/recipes/librispeech/index.html> for how to run models in this recipe.
|
|
|
|
[./RESULTS.md](./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 |
|
|
|---------------------------------------|---------------------|--------------------|---------------------------------------------------|
|
|
| `transducer` | Conformer | LSTM | |
|
|
| `transducer_stateless` | Conformer | Embedding + Conv1d | Using optimized_transducer from computing RNN-T loss |
|
|
| `transducer_stateless2` | Conformer | Embedding + Conv1d | Using torchaudio for computing RNN-T loss |
|
|
| `transducer_lstm` | LSTM | LSTM | |
|
|
| `transducer_stateless_multi_datasets` | Conformer | Embedding + Conv1d | Using data from GigaSpeech as extra training data |
|
|
| `pruned_transducer_stateless` | Conformer | Embedding + Conv1d | Using k2 pruned RNN-T loss |
|
|
| `pruned_transducer_stateless2` | Conformer(modified) | Embedding + Conv1d | Using k2 pruned RNN-T loss |
|
|
| `pruned_transducer_stateless3` | Conformer(modified) | Embedding + Conv1d | Using k2 pruned RNN-T loss + using GigaSpeech as extra training data |
|
|
| `pruned_transducer_stateless4` | Conformer(modified) | Embedding + Conv1d | same as pruned_transducer_stateless2 + save averaged models periodically during training |
|
|
| `pruned_transducer_stateless5` | Conformer(modified) | Embedding + Conv1d | same as pruned_transducer_stateless4 + more layers + random combiner|
|
|
| `pruned_transducer_stateless6` | Conformer(modified) | Embedding + Conv1d | same as pruned_transducer_stateless4 + distillation with hubert|
|
|
|
|
|
|
The decoder in `transducer_stateless` is modified from the paper
|
|
[Rnn-Transducer with Stateless Prediction Network](https://ieeexplore.ieee.org/document/9054419/).
|
|
We place an additional Conv1d layer right after the input embedding layer.
|