syang1993/gst-tacotron

Issues

• training stops many seconds to create new queue of data

  #50 opened 3 years ago by Adibian
  0

• shape of linear_outputs is not same as while training

  #49 opened 3 years ago by Mihir-Gajera1
  0

• Add style weights when there is no reference audio

  #48 opened 3 years ago by zh1yu4nyu
  0

• Using pre-trained model of Keithito's tacotron implementation

  #47 opened 3 years ago by hariduttt
  0

• Regarding the trained model

  #46 opened 3 years ago by hariduttt
  0

• eval form checkpoints

  #4 opened 7 years ago by marymirzaei
  30

• Training with custom data

  #23 opened 6 years ago by wanshun123
  2

• Unable to reproduce results

  #44 opened 4 years ago by Anchit1999
  0

• Pretrained Model

  #26 opened 6 years ago by harismuneer
  3

• Mumbling in synthesis

  #45 opened 4 years ago by a-froghyar
  1

• Style Token Layer implementation question

  #1 opened 7 years ago by acetylSv
  20

• Please update link to Blizzard data

  #28 opened 6 years ago by simonkingedinburgh
  4

• How to achieve style embedding with different weights of each token without reference audio?

  #29 opened 6 years ago by bitwangyujia
  2

• Pretrained Weights

  #43 opened 5 years ago by ashish-roopan
  1

• preprocessing the training data

  #2 opened 7 years ago by marymirzaei
  2

• What is in reference audio path?

  #42 opened 5 years ago by Thien223
  0

• can we synthesis speaker-A's tone with speaker-B's prosody?

  #41 opened 5 years ago by niu0717
  0

• Path for Reference Audio

  #38 opened 5 years ago by shrinidhin
  1

• erro in eval.py

  #39 opened 5 years ago by 1105060120
  1

• Check failed: dnnReLUCreateBackward_F32

  #40 opened 5 years ago by miyoungvkim
  1

• Error in datafeeder.py

  #37 opened 5 years ago by shrinidhin
  1

• poor alignment when conditioned on reference audios

  #20 opened 6 years ago by mohsinjuni
  4

• Why use the 'tf.layer.conv1d' for query, key transformation instead of fully connected layer?

  #36 opened 5 years ago by LEEYOONHYUNG
  0

• where do you insert or import wav file of models voice for training?

  #35 opened 5 years ago by pnwseeker
  0

• Reference Encoder Padding

  #34 opened 6 years ago by its-sandy
  0

• Some problems when preprocessing ljspeech dataset

  #33 opened 6 years ago by Charlottecuc
  1

• Training Multi-Speaker Model.

  #21 opened 6 years ago by sujithpadar
  5

• Throws "data must be floating-point" exception after 1k steps

  #25 opened 6 years ago by ishandutta2007
  12

• No clear speech

  #32 opened 6 years ago by ErnstTmp
  7

• GMM Attention

  #31 opened 6 years ago by ErnstTmp
  5

• tensorflow.python.framework.errors_impl.InvalidArgumentError: Incompatible shapes: [12,2262,80] vs. [12,2000,80]

  #30 opened 6 years ago by ErnstTmp
  4

• how long will I wait for the Blizzard dataset confirmation?

  #27 opened 6 years ago by XiaooaiX
  5

• Why there is some blank in the sythesized wav file when we use reference audio generation?

  #24 opened 6 years ago by begeekmyfriend
  0

• core dumped error when preprocessing ljspeech dataset

  #22 opened 6 years ago by wanshun123
  2

• Sample Alignment Graph

  #10 opened 6 years ago by fazlekarim
  4

• Train as a Tacotron1 script problem

  #15 opened 6 years ago by dazenhom
  4

• poor alignment when synthesizing long sentences

  #19 opened 6 years ago by moonnee
  1

• Tone transfer

  #13 opened 6 years ago by switchzts
  4

• the model is hard to converge with LJSpeech

  #18 opened 6 years ago by zyj008
  1

• poor alignment with test out-of-collection data

  #16 opened 6 years ago by butterl
  3

• How to integrate this code to r9y9's wavenet_vocoder ?

  #14 opened 6 years ago by rishikksh20
  13

• Eval on soft voices

  #17 opened 6 years ago by fazlekarim
  0

• multi head attention

  #12 opened 6 years ago by Young-Sun
  2

• Preprocessing blizzard 2013 data

  #11 opened 6 years ago by jsonko
  2

• training time

  #8 opened 7 years ago by Young-Sun
  4

• data feeder error

  #6 opened 7 years ago by fazlekarim
  6

• What would happen if we merged datasets?

  #5 opened 7 years ago by fazlekarim
  3