mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-08-08 17:42:21 +00:00
Code Issues Packages Projects Releases Wiki Activity
icefall/egs/librispeech/ASR
History
Mingshuang Luo 8cb727e24a
Tedlium3 pruned transducer stateless (#261) 
* update tedlium3-pruned-transducer-stateless-codes

* update README.md

* update README.md

* add fast beam search for decoding

* do a change for RESULTS.md

* do a change for RESULTS.md

* do a fix

* do some changes for pruned RNN-T
2022-04-11 17:08:53 +08:00
..
conformer_ctc
Fix potential bugs in PyTorch that exist in label_smoothing. (#300)
2022-04-08 13:41:33 +08:00
conformer_mmi
Change for asr_datamodule.py (#241)
2022-03-14 00:30:58 +08:00
local
Begin to use multiple datasets in training (#213)
2022-02-21 15:27:27 +08:00
pruned_transducer_stateless
Tedlium3 pruned transducer stateless (#261)
2022-04-11 17:08:53 +08:00
pruned_transducer_stateless2
Support mix precision training on the reworked model (#305)
2022-04-11 16:49:54 +08:00
streaming_conformer_ctc
Reset seed at the beginning of each epoch. (#221)
2022-02-21 15:16:39 +08:00
tdnn_lstm_ctc
Some cleanups
2022-04-04 13:37:10 +08:00
transducer
Set the seed for dataloader. (#282)
2022-03-31 16:48:46 +08:00
transducer_lstm
Set the seed for dataloader. (#282)
2022-03-31 16:48:46 +08:00
transducer_stateless
Some cleanups
2022-04-04 13:37:10 +08:00
transducer_stateless_multi_datasets
Don't use a lambda for dataloader's worker_init_fn. (#284)
2022-03-31 20:32:00 +08:00
prepare_giga_speech.sh
Begin to use multiple datasets in training (#213)
2022-02-21 15:27:27 +08:00
prepare.sh
Add force alignment for stateless transducer. (#239)
2022-03-12 16:16:15 +08:00
README.md
Add modified beam search for pruned rnn-t. (#248)
2022-03-12 16:16:55 +08:00
RESULTS-100hours.md
Update result for full libri + GigaSpeech using transducer_stateless. (#231)
2022-03-01 17:01:46 +08:00
RESULTS.md
Add modified beam search for pruned rnn-t. (#248)
2022-03-12 16:16:55 +08:00
shared
Refactoring (#4)
2021-08-04 14:53:02 +08:00

README.md

Introduction

Please refer to https://icefall.readthedocs.io/en/latest/recipes/librispeech/index.html for how to run models in this recipe.

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
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

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.