barronalex/Tacotron

Issues

• Training time

  #34 opened 3 years ago by urmikakasi
  0

• help with the output file

  #33 opened 4 years ago by swarajbadhei
  0

• General Question

  #32 opened 5 years ago by ShaharDagan
  0

• Inference error

  #31 opened 5 years ago by kaushikb258
  0

• Attention RNN

  #30 opened 6 years ago by pravn
  1

• maximum_audio_length in audio.py, what is it and what are its units?

  #29 opened 6 years ago by egaebel
  0

• weights.zip download issue

  #28 opened 7 years ago by ahmednabil950
  0

• French Model

  #25 opened 7 years ago by dallal9
  0

• Any way to freeze the model?

  #24 opened 7 years ago by dreamibor
  0

• Hyperparameter estimation

  #23 opened 7 years ago by ErfolgreichCharismatisch
  0

• MultiSpeaker test.py

  #22 opened 7 years ago by gaziway
  0

• Anyone getting good results on the multi-speaker VCTK dataset?

  #21 opened 7 years ago by gaziway
  0

• Another reason to go multi speaker now

  #18 opened 7 years ago by GunpowderGuy
  1

• Terrible result at 180k steps with arctic dataset

  #20 opened 7 years ago by tuong-olli
  0

• Annealing rate question

  #19 opened 7 years ago by thuoctran
  0

• Anyone getting good results on the nancy dataset?

  #12 opened 7 years ago by jacksmite
  1

• different results output during training compared to test.py

  #15 opened 7 years ago by mrelich
  19

• loss exploded

  #7 opened 7 years ago by YisenWang
  3

• Pre-trained model?

  #1 opened 7 years ago by greigs
  1

• High Memory Usage?

  #6 opened 7 years ago by zhang-jian
  4

• don't get anything by running test.py, how to generate the wav?

  #14 opened 7 years ago by azraelkuan
  2

• promising results

  #4 opened 8 years ago by zhang-jian
  8

• test.py arctic dataset out of range

  #13 opened 7 years ago
  0

• Samples generated in train.py

  #11 opened 8 years ago by jacksmite
  1

• Test won't generate speech used in training while training does generate correctly.

  #9 opened 8 years ago by onyedikilo
  3

• Utf-8 characters for non English words.

  #10 opened 8 years ago by onyedikilo
  1

• No download file http://data.cstr.ed.ac.uk

  #8 opened 8 years ago by ytgcljj
  0

• Why not use the mean of the loss?

  #5 opened 8 years ago by auzxb
  0

• Better datasets to consider

  #2 opened 8 years ago by greigs
  0