Guoguo Chen
|
0edb20ddb3
|
mpe training parameters for Babel fullLP; limitedLP needs re-investigation as we didn't run that after fixing the normalization stuff.
git-svn-id: https://svn.code.sf.net/p/kaldi/code/trunk@3481 5e6a8d80-dfce-4ca6-a32a-6e07a63d50c8
|
2014-02-07 20:33:21 +00:00 |
Dan Povey
|
25183ff180
|
trunk: BABEL scripts; Mostly cosmetic changes and minor fixes to scripts.
git-svn-id: https://svn.code.sf.net/p/kaldi/code/trunk@3455 5e6a8d80-dfce-4ca6-a32a-6e07a63d50c8
|
2014-02-01 23:58:12 +00:00 |
|
Guoguo Chen
|
b71ec1af8e
|
Fix to discriminative training (retroactive thing and nnet-normalize-stddev), seems working, learning rate needs further tuning
git-svn-id: https://svn.code.sf.net/p/kaldi/code/trunk@3421 5e6a8d80-dfce-4ca6-a32a-6e07a63d50c8
|
2014-01-20 19:12:28 +00:00 |
|
Guoguo Chen
|
2cafa2ab33
|
scripts for discriminative training on top of pnorm recipe
git-svn-id: https://svn.code.sf.net/p/kaldi/code/trunk@3374 5e6a8d80-dfce-4ca6-a32a-6e07a63d50c8
|
2013-12-29 17:46:40 +00:00 |