Fangjun Kuang
|
fc7574f6d2
|
Add results.
|
2022-04-29 12:03:22 +08:00 |
|
Fangjun Kuang
|
187534df2e
|
Merge branch 'modified-conformer-with-multi-datasets' of github.com:csukuangfj/icefall into modified-conformer-with-multi-datasets
|
2022-04-28 06:54:56 +08:00 |
|
Fangjun Kuang
|
af20922320
|
Minor fixes.
|
2022-04-28 06:51:11 +08:00 |
|
Fangjun Kuang
|
b1c3705fbe
|
Compute the Nbest oracle WER for RNN-T decoding.
|
2022-04-24 15:10:30 +08:00 |
|
Fangjun Kuang
|
a31207f5b3
|
Fix decoding the gigaspeech dataset.
We have to use the decoder/joiner networks for the GigaSpeech dataset.
|
2022-04-18 15:23:07 +08:00 |
|
Fangjun Kuang
|
e32641d1df
|
Add decoding code for GigaSpeech.
|
2022-04-18 13:27:53 +08:00 |
|
Fangjun Kuang
|
5c7c9918a4
|
Fix warnings in decoding.
|
2022-04-17 22:59:19 +08:00 |
|
Fangjun Kuang
|
f0330f9d2d
|
Combine XL splits lazily during training.
|
2022-04-17 21:45:55 +08:00 |
|
Fangjun Kuang
|
0cc13bc702
|
Support specifying number of workers for on-the-fly feature extraction.
|
2022-04-14 10:26:45 +08:00 |
|
Fangjun Kuang
|
bbf074a36b
|
Use librispeech + gigaspeech with modified conformer.
|
2022-04-12 17:28:01 +08:00 |
|
Fangjun Kuang
|
337309267b
|
Copy files for editing.
|
2022-04-12 16:16:40 +08:00 |
|