mirrors/icefall
1
0
Fork 0
mirror of https://github.com/k2-fsa/icefall.git synced 2025-08-27 02:34:21 +00:00
Code Issues Packages Projects Releases Wiki Activity
102 Commits 5 Branches 3 Tags
Commit Graph

102 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
Mingshuang Luo
b35eed961a
Merge pull request #1 from luomingshuang/lossrecord-v2 
style check with flake8 and black
 2021-09-29 19:29:25 +08:00
Mingshuang Luo
79fd09e3e5
Add files via upload 2021-09-29 19:24:40 +08:00
Mingshuang Luo
279dc74b4e
Add files via upload 2021-09-29 19:23:54 +08:00
Mingshuang Luo
bdd890bab9
Add files via upload 2021-09-29 19:23:11 +08:00
Mingshuang Luo
0dfe0e6680
Add files via upload 2021-09-29 19:22:36 +08:00
Mingshuang Luo
05a79c2cbb
Delete utils.py 2021-09-29 19:22:16 +08:00
Mingshuang Luo
f531615d93
Add files via upload 2021-09-29 19:12:47 +08:00
Mingshuang Luo
988d7af85c
Delete utils.py 2021-09-29 19:12:02 +08:00
Mingshuang Luo
f837818af3
Add files via upload 2021-09-29 19:11:14 +08:00
Mingshuang Luo
cca1399e0f
Delete train.py 2021-09-29 19:10:49 +08:00
Mingshuang Luo
67762e308f
Add files via upload 2021-09-29 19:10:33 +08:00
Mingshuang Luo
49af342863
Delete train.py 2021-09-29 19:10:08 +08:00
Mingshuang Luo
7278b699b3
Add files via upload 2021-09-29 19:09:36 +08:00
Mingshuang Luo
3c5b49006e
Delete train.py 2021-09-29 19:09:11 +08:00
Mingshuang Luo
852efe1b87
Rename train_lossrecord.py to train.py 2021-09-29 17:37:21 +08:00
Mingshuang Luo
8b6c139623
Add files via upload 2021-09-29 17:36:52 +08:00
Mingshuang Luo
bc101093c6
Delete train.py 2021-09-29 17:36:30 +08:00
Mingshuang Luo
00b2d4c9c7
Rename train_lossrecord_tdnn.py to train.py 2021-09-29 17:35:58 +08:00
Mingshuang Luo
84eb064675
Add files via upload 2021-09-29 17:35:24 +08:00
Mingshuang Luo
f03fb67aa1
Delete train.py 2021-09-29 17:35:03 +08:00
Mingshuang Luo
426418c703
Rename train_lossrecord_con.py to train.py 2021-09-29 17:34:34 +08:00
Mingshuang Luo
787ca3b89c
Add files via upload 2021-09-29 17:33:56 +08:00
Mingshuang Luo
6de7f0c062
Delete train.py 2021-09-29 17:33:32 +08:00
Mingshuang Luo
7c1819dddd
Add files via upload 2021-09-29 17:33:03 +08:00
Mingshuang Luo
94ef8208b1
Delete utils.py 2021-09-29 17:32:34 +08:00
Mingshuang Luo
de9b2a9cd1
Update train.py 2021-09-29 16:54:21 +08:00
Mingshuang Luo
a0994ee58a
Update utils.py 2021-09-29 16:48:00 +08:00
Mingshuang Luo
7faab70ddf
Update utils.py 2021-09-29 16:43:01 +08:00
Mingshuang Luo
ec3b1bb015
Update utils.py 2021-09-29 16:11:23 +08:00
Mingshuang Luo
d0986e9787
Update utils.py 2021-09-29 16:04:56 +08:00
Mingshuang Luo
d04271219c
Update utils.py 2021-09-29 16:01:53 +08:00
Mingshuang Luo
28fa30a255
Update utils.py 2021-09-29 14:13:31 +08:00
Mingshuang Luo
8d07ce2185
Update utils.py 2021-09-29 13:11:52 +08:00
Mingshuang Luo
1c0792796b
Update train.py 2021-09-29 12:57:23 +08:00
Mingshuang Luo
43cf016ae5
Update train.py 2021-09-29 12:56:47 +08:00
Mingshuang Luo
34e36a926b
Update train.py 2021-09-29 12:55:42 +08:00
Mingshuang Luo
597ff01158
Update train.py 2021-09-29 12:51:38 +08:00
Mingshuang Luo
0fa46bf68a
Update train.py 2021-09-29 12:49:59 +08:00
Mingshuang Luo
e74e75acc6 Use LossRecord to record and print loss for the training process 2021-09-29 10:08:38 +08:00
Mingshuang Luo
73f21a379b
Merge branch 'k2-fsa:master' into master 2021-09-27 15:34:08 +08:00
Fangjun Kuang
707d7017a7
Support pure ctc decoding requiring neither a lexicon nor an n-gram LM (#58) 
* Rename lattice_score_scale to nbest_scale.

* Support pure CTC decoding requiring neither a lexicion nor an n-gram LM.

* Fix style issues.

* Fix a typo.

* Minor fixes.
 2021-09-26 14:21:49 +08:00
Mingshuang Luo
6c4a58273f
Fix some spelling errors. 2021-09-26 12:55:51 +08:00
Mingshuang Luo
6abd1bcd0a
Fix some spelling errors. 2021-09-26 12:54:35 +08:00
Mingshuang Luo
662690aa8b
Merge branch 'k2-fsa:master' into master 2021-09-26 12:51:30 +08:00
Fangjun Kuang
455693aede
Fix hasattr of AttributeDict. (#52) 2021-09-22 16:37:20 +08:00
Fangjun Kuang
a80e58e15d
Refactor decode.py to make it more readable and more modular. (#44) 
* Refactor decode.py to make it more readable and more modular.

* Fix an error.

Nbest.fsa should always have token IDs as labels and
word IDs as aux_labels.

* Add nbest decoding.

* Compute edit distance with k2.

* Refactor nbest-oracle.

* Add rescore with nbest lists.

* Add whole-lattice rescoring.

* Add rescoring with attention decoder.

* Refactoring.

* Fixes after refactoring.

* Fix a typo.

* Minor fixes.

* Replace [] with () for shapes.

* Use k2 v1.9

* Use Levenshtein graphs/alignment from k2 v1.9

* [doc] Require k2 >= v1.9

* Minor fixes.
 2021-09-20 15:44:54 +08:00
Fangjun Kuang
cc77cb3459
Fix decode.py to remove the correct axis. (#50) 
* Fix decode.py to remove the correct axis.

* Run GitHub actions manually.
 2021-09-17 16:49:03 +08:00
Wei Kang
9a6e0489c8
update api for RaggedTensor (#45) 
* Fix code style

* update k2 version in CI

* fix compile hlg
 2021-09-14 16:39:56 +08:00
Fangjun Kuang
a2be2896a9
Fix the link to k2's installation doc. (#46) 2021-09-14 13:39:52 +08:00
Wei Kang
24656e9749
Update docs and remove unnecessary arguments (#42) 
* Fix typo in docs

* Update docs and remove unnecessary arguments

* Fix code style
 2021-09-13 18:28:57 +08:00