malarinv
/
tacotron2
mirror of https://github.com/malarinv/tacotron2
1
0
Fork 0
Code Issues Releases Wiki Activity
108 Commits
10 Branches
0 Tags
3.3 MiB
Commit Graph

108 Commits (d6670c8ed70d497dea854173061730e1514b3a96)

Author SHA1 Message Date
rafaelvalle d6670c8ed7 Dockerfile: updating to use latest pytorch and apex 2019-04-03 13:51:22 -08:00
rafaelvalle 0274619e45 train.py: using amp for mixed precision training 2019-04-03 13:42:00 -08:00
rafaelvalle bb20035586 inference.ipynb: adding fp16 inference 2019-04-03 13:41:11 -08:00
rafaelvalle 1480f82908 model.py: renaming variables, removing dropout from lstm cell state, removing conversions now handled by amp 2019-04-03 13:36:35 -08:00
rafaelvalle 087c86755f logger.py: using new pytorch api 2019-04-03 13:35:04 -08:00
rafaelvalle f37998c59d train.py: shuffling at every epoch 2019-03-15 17:49:27 -07:00
rafaelvalle bff304f432 README.md: adding explanation on training from pre-trained model 2019-03-15 17:38:40 -07:00
rafaelvalle 3869781877 train.py: adding routine to warm start and ignore layers, e.g. embedding.weight 2019-03-15 17:34:27 -07:00
rafaelvalle bb67613493 hparams.py: adding ignore_layers argument to ignore text embedding layers when warm_starting 2019-03-15 17:28:50 -07:00
rafaelvalle af1f71a975 inference.ipynb: adding code to remove waveglows bias 2019-03-15 16:54:54 -07:00
rafaelvalle fc0d34cfce stft.py: moving window_sum to cuda if magnitude is cuda 2019-03-15 14:36:56 -07:00
Rafael Valle f2c94d94fd
Merge pull request #136 from GrzegorzKarchNV/master 
Fixing concatenation error for fp16 distributed training
 2019-02-01 12:10:42 -08:00
gkarch df4a466af2 Fixing concatenation error for fp16 ditributed training 2019-02-01 09:55:59 +01:00
rafaelvalle 825ffa47d1 inference.ipynb: reverting fp16 inference for now 2018-12-08 21:26:01 -08:00
rafaelvalle 4d7b04120a inference.ipynb: changing waverglow inference fo fp16 2018-12-05 22:14:35 -08:00
rafaelvalle 6e430556bd train.py: val logger on gpu 0 only 2018-11-27 22:03:11 -08:00
rafaelvalle 3973b3e495 hparams.py: distributed using tcp 2018-11-27 22:02:43 -08:00
rafaelvalle 52a30bb7b6 distributed.py: replacing to avoid distributed error 2018-11-27 21:01:26 -08:00
rafaelvalle 0ad65cc053 train.py: renaming variable to n_gpus 2018-11-27 21:00:05 -08:00
rafaelvalle 8300844fa7 hparams.py: removing 22khz 2018-11-27 20:56:52 -08:00
rafaelvalle f06063f746 train.py: renaming function, removing dataparallel 2018-11-27 18:04:12 -08:00
rafaelvalle 3045ba125b inference.ipynb: cleanup 2018-11-27 12:04:36 -08:00
rafaelvalle 4c4aca3662 README.md: layout 2018-11-27 11:59:05 -08:00
rafaelvalle 05dd8f91d2 README.md: adding submodule init to README 2018-11-27 11:55:40 -08:00
rafaelvalle 5d66c3deab adding waveglow submodule 2018-11-27 11:53:20 -08:00
Rafael Valle f02704f338
Merge pull request #96 from NVIDIA/clean_slate 
Clean slate
 2018-11-27 08:06:00 -08:00
rafaelvalle ba8cf36198 requirements.txt: removing pytorch 0.4 from requirements. upgrading to 1.0 2018-11-27 08:04:21 -08:00
rafaelvalle b5e0a93946 inference.ipynb: updating inference file with relative paths 2018-11-27 08:04:04 -08:00
rafaelvalle 58b0ec61bd README.md: updating requirements and inference demo 2018-11-27 08:03:34 -08:00
rafaelvalle 1ad939df1a inference.ipynb: setting relative model paths 2018-11-27 07:45:09 -08:00
rafaelvalle 32b9a135d0 utils.py: updating 2018-11-25 22:34:38 -08:00
rafaelvalle ce29e13959 train.py: updating 2018-11-25 22:34:34 -08:00
rafaelvalle 1ea6ed5861 text/symbols.py: updating symbols 2018-11-25 22:34:26 -08:00
rafaelvalle cdfde985e5 text/__init__.py: remove stop token 2018-11-25 22:34:11 -08:00
rafaelvalle e314bb4cd0 stft.py: fix filter winlength error 2018-11-25 22:33:52 -08:00
rafaelvalle 4af4ccb135 model.py: rewrite 2018-11-25 22:33:38 -08:00
rafaelvalle 1ec0e5e8cd layers.py: rewrite 2018-11-25 22:33:32 -08:00
rafaelvalle 249afd8043 inference.ipynb: import taco2model to be public 2018-11-25 22:33:16 -08:00
rafaelvalle 1b243d5d5a hparams.py:rewrite 2018-11-25 22:33:05 -08:00
rafaelvalle d0aa9e7d32 distributed.py: rewrite 2018-11-25 22:32:54 -08:00
rafaelvalle 1683a57ae5 data_utils.py: rewrite 2018-11-25 22:32:47 -08:00
Rafael Valle fc0cf6a89a
Merge pull request #53 from cobr123/patch-1 
add pillow
 2018-07-02 14:32:21 -07:00
cobr123 8de38495be
add pillow 2018-07-02 19:35:51 +03:00
rafaelvalle 7eb045206c README.md: updating readme to explicitly mention that mel representation of WaveNet and Tacotron2 must be the same 2018-06-14 11:25:42 -07:00
rafaelvalle c67005f1be Dockerfile: adding jupyter to dockerfile 2018-06-14 10:30:01 -07:00
rafaelvalle cb3794796f Dockerfile: removing return from Dockerfile 2018-06-12 21:38:03 -07:00
Rafael Valle a8de973923
Merge pull request #37 from yoks/master 
`used_saved_learning_rate`  fix
 2018-06-12 09:00:14 -07:00
yoks a0ae2da05f
`used_saved_learning_rate` fix 
`used_saved_learning_rate` name change to `use_saved_learning_rate`
 2018-06-12 16:53:12 +03:00
rafaelvalle 34066ac4fc requirements.txt: setting torch to 0.4.0 2018-06-11 08:16:31 -07:00
rafaelvalle 12ab5ba89c model.py: setting weight initialization to xavier uniform 2018-06-07 20:28:52 -07:00