coqui-ai
/
TTS
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
1,346 Commits
80 Branches
98 Tags
316 MiB
Commit Graph

1346 Commits (263dc2f7ce067ad5535f779f9ff2d0bad2cb4a2f)

Author SHA1 Message Date
erogol 263dc2f7ce fix weird lint problem but not caring 2020-05-20 16:27:48 +02:00
erogol ddd7de6439 update notebook 2020-05-20 16:12:10 +02:00
erogol 4a6949632b bug fixes and benchmark notebook update 2020-05-20 16:05:57 +02:00
erogol ed67cadf98 requirements for testingwq 2020-05-20 15:06:41 +02:00
erogol a893405739 raise not implemented for multispeaker TTS_tf inference 2020-05-20 14:26:56 +02:00
erogol cb9ac27b65 TTS_tf notebook update 2020-05-20 14:26:47 +02:00
erogol 6ccf32c2b9 update tests 2020-05-20 14:00:31 +02:00
erogol ca359727bc config update and change default debug mode 2020-05-20 12:30:06 +02:00
erogol 1835628335 tf conversion fixes 2020-05-20 12:25:24 +02:00
erogol dc166b42e3 update config.json 2020-05-20 11:55:32 +02:00
erogol 97cd39bf99 console logger fix 2020-05-19 16:44:37 +02:00
erogol f75b0a6439 linter updates 2020-05-18 18:46:13 +02:00
erogol 496ff68dec config update 2020-05-18 18:45:30 +02:00
erogol 327c88b4bb dme update 2020-05-18 13:31:14 +02:00
erogol df8fd3823d Merge branch 'tf-convert2' into dev 2020-05-18 13:13:21 +02:00
Eren Gölge e55e28bb37 Merge branch 'dev' of https://github.com/mozilla/TTS into dev 2020-05-18 13:12:31 +02:00
erogol 342d6303d4 update TF model notebook 2020-05-18 12:20:51 +02:00
erogol 8e6aedccee update readme 2020-05-18 12:00:10 +02:00
erogol 523fa5dfd2 pass sequence mask to the same device as the input 2020-05-18 11:35:19 +02:00
erogol 8805370645 add tf tacotron2 test and edit test utils imports after utils 
refactoring
 2020-05-18 11:34:13 +02:00
Eren Gölge 67397be1c0 tf folder add 2020-05-18 11:02:36 +02:00
Eren Gölge d023872f0d
Merge pull request #407 from thllwg/dev 
Configurable number of workers for DataLoader
 2020-05-16 15:45:07 +02:00
Eren Gölge 93cab61d80
Merge pull request #406 from mittimithai/patch-6 
small change for multispeaker
 2020-05-15 11:11:04 +02:00
thllwg 65b9c7d3d6 number of workers as config parameter 2020-05-15 10:19:52 +02:00
erogol d5d9e6e8ea bug fix 2020-05-13 13:52:17 +02:00
mittimithai 25f466f299
more whitespace problems 2020-05-12 19:02:37 -07:00
mittimithai 42ff83f9b9
trying to fix trailing whitespace 2020-05-12 18:50:58 -07:00
mittimithai a4aca623c3
removed + chars 
silly mistake copy pasting
 2020-05-12 15:23:45 -07:00
mittimithai 85a822e319
small change for multispeaker 
just threads speaker_id through decoder.run_model
 2020-05-12 15:02:24 -07:00
erogol 1cd25ccf0d bug fix 2020-05-12 16:23:32 +02:00
erogol 68dbcee746 import condition update for synthesis with TF 2020-05-12 16:23:32 +02:00
erogol 84c5c4a587 config remove empty chars 2020-05-12 16:23:32 +02:00
erogol b3ec50b5c4 tf bacend for synthesis 2020-05-12 16:23:32 +02:00
erogol d99fda8e42 init batch norm explicit initial values 2020-05-12 16:23:32 +02:00
erogol 6f5c8773d6 enable encoder lstm bias 2020-05-12 16:23:32 +02:00
erogol 9504b71f79 fix lstm biases True 2020-05-12 16:23:32 +02:00
erogol de2918c85b bug fixes 2020-05-12 16:23:32 +02:00
erogol 736f169cc9 tf lstm does not match torch lstm wrt bias vectors. So I avoid bias in LSTM as an easy solution. 2020-05-12 16:23:32 +02:00
erogol d282222553 renaming layers to be converted to TF counterpart 2020-05-12 16:23:32 +02:00
erogol bee288fa93 fixing console logging colors 2020-05-12 16:22:59 +02:00
erogol 88bde77061 fix checkpointing 2020-05-12 14:09:28 +02:00
erogol 0ec42fa279 more agressive remove folder 2020-05-12 14:09:16 +02:00
erogol 3b2d726e2d radam pytorch 1.5 update 2020-05-12 13:58:10 +02:00
erogol 574968b249 refactoring utils 2020-05-12 13:57:37 +02:00
erogol 720c4690db update imports 2020-05-12 13:57:26 +02:00
erogol c0c3c6e331 train.py update imports for utils refactoring 2020-05-12 13:57:13 +02:00
erogol 2d9dcd60ba update imports for util refactoring 2020-05-12 13:56:49 +02:00
Eren Gölge 7292d303b9
Merge pull request #404 from Edresson/patch-3 
Fix bug in Graves Attn
 2020-05-07 17:37:25 +02:00
Eren Gölge 7a9247f91f
Merge pull request #403 from Edresson/patch-2 
fix bug in bidirectional decoder train
 2020-05-05 00:33:53 +02:00
Edresson Casanova cce13ee245
Fix bug in Graves Attn 
On my machine at Graves attention the variable self.J ( self.J = torch.arange(0, inputs.shape[1]+2).to(inputs.device) + 0.5) is a LongTensor, but it must be a float tensor. So I get the following error:

Traceback (most recent call last):
  File "train.py", line 704, in <module>
    main(args)
  File "train.py", line 619, in main
    global_step, epoch)
  File "train.py", line 170, in train
    text_input, text_lengths, mel_input, speaker_embeddings=speaker_embeddings)
  File "/home/edresson/anaconda3/envs/TTS2/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/mnt/edresson/DD/TTS/voice-clonning/TTS/tts_namespace/TTS/models/tacotron.py", line 121, in forward
    self.speaker_embeddings_projected)
  File "/home/edresson/anaconda3/envs/TTS2/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/mnt/edresson/DD/TTS/voice-clonning/TTS/tts_namespace/TTS/layers/tacotron.py", line 435, in forward
    output, stop_token, attention = self.decode(inputs, mask)
  File "/mnt/edresson/DD/TTS/voice-clonning/TTS/tts_namespace/TTS/layers/tacotron.py", line 367, in decode
    self.attention_rnn_hidden, inputs, self.processed_inputs, mask)
  File "/home/edresson/anaconda3/envs/TTS2/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/mnt/edresson/DD/TTS/voice-clonning/TTS/tts_namespace/TTS/layers/common_layers.py", line 180, in forward
    phi_t = g_t.unsqueeze(-1) * (1.0 / (1.0 + torch.sigmoid((mu_t.unsqueeze(-1) - j) / sig_t.unsqueeze(-1))))
RuntimeError: expected type torch.cuda.FloatTensor but got torch.cuda.LongTensor


In addition the + 0.5 operation is canceled if it is a LongTensor.
Test: 
>>> torch.arange(0, 10) 
tensor([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> torch.arange(0, 10) + 0.5
tensor([0, 1, 2, 3, 4, 5, 6, 7, 8, 9])
>>> torch.arange(0, 10.0) + 0.5
tensor([0.5000, 1.5000, 2.5000, 3.5000, 4.5000, 5.5000, 6.5000, 7.5000, 8.5000,
        9.5000])

To resolve this I forced the arrange range to float:
self.J = torch.arange(0, inputs.shape[1]+2.0).to(inputs.device) + 0.5
 2020-05-04 17:52:58 -03:00