p0p4k/vits2_pytorch

Issues

• When we should use duration discriminator ?

  #104 opened a month ago by Marioando
  0

• about duration-discriminator training objective

  #59 opened a year ago by choiHkk
  9

• Training fails C:\actions-runner\_work\pytorch\pytorch\builder\windows\pytorch\aten\src\ATen\native\cuda\Indexing.cu:1289: block: [523,0,0], thread: [31,0,0] Assertion `srcIndex < srcSelectDimSize` failed.

  #103 opened 2 months ago by SeymourNickelson
  5

• attentions.Decoder, what is this used for?

  #100 opened 3 months ago by hildazzz
  1

• 请问，我数据处理生成音素的时候出现这个问题，是什么原因呢

  #98 opened 4 months ago by pengpengtao
  0

• 你好，有中文的模型下载吗

  #94 opened 6 months ago by TszSimLaw
  1

• AlignerNet instead of MAS

  #81 opened 10 months ago by codeghees
  18

• RuntimeError: The expanded size of the tensor

  #92 opened 7 months ago by Moon-sung-woo
  4

• RuntimeError: during training

  #91 opened 7 months ago by ArthurMalajyan
  0

• Why are you using "use_mel_posterior_encoder"?

  #89 opened 8 months ago by Moon-sung-woo
  4

• After finishing the training, continuing the training revealed that the curve of learning rate had reversed。

  #88 opened 8 months ago by Mr47121836
  6

• Training stuck

  #71 opened a year ago by Madhavan0123
  18

• why 'duration_discriminator_type' is redefined?

  #84 opened 9 months ago by 5Hyeons
  0

• Training using SDP (and with DP by ratio?)

  #79 opened 10 months ago by isdanni
  6

• Have anyone tried to decrease the feature channels of the model?

  #80 opened 10 months ago by JohnHerry
  0

• I am trying to use the vits's New FLOW in the svc

  #22 opened a year ago by Hanxibird
  15

• good, how this combine to the bert-vits2?

  #75 opened 10 months ago by foocker
  2

• hop size issue

  #78 opened a year ago by Paintako
  1

• Keep training on existed checkpoint

  #76 opened a year ago by HuuHuy227
  3

• Training doesn´t start when speaker IDs isn´t sequential from 0

  #77 opened a year ago by skol101
  1

• Bad quality audio when infer with custom condition

  #74 opened a year ago by huutuongtu
  1

• colab error

  #73 opened a year ago by alexxony
  3

• Where is L_mse?

  #72 opened a year ago by OedoSoldier
  11

• how to know whether the model is fitting

  #66 opened a year ago by creed2415
  1

• Duration Discriminator problem

  #68 opened a year ago by yiwei0730
  18

• Export to JIT script

  #46 opened a year ago by OnceJune
  4

• DurationDiscriminatorType error

  #64 opened a year ago by lpscr
  5

• New feature: Deleting the old .pth files when training

  #56 opened a year ago by KevinWang676
  3

• Checkpoint saves?

  #58 opened a year ago by ziyaad30
  1

• Just curious, for line 528 and line 530 in models.py, why return x without calculation?

  #45 opened a year ago by hildazzz
  1

• ValueError: too many values to unpack (expected 2)

  #57 opened a year ago by KevinWang676
  5

• KeyError when training the Chinese dataset

  #53 opened a year ago by KevinWang676
  4

• KSS dataset for 490 epochs but the quality is not as good as I expected

  #49 opened a year ago by ggpid
  4

• will more ckpts will be add later?

  #52 opened a year ago by amssss0
  0

• more checkpoints.

  #50 opened a year ago by amssss0
  1

• vctk dataset training reproduction of repo

  #47 opened a year ago by I8dNLo
  5

• Fine tuning

  #40 opened a year ago by CCludts
  1

• Error when training model: AttributeError: '_MultiProcessingDataLoaderIter' object has no attribute '_shutdown'

  #33 opened a year ago by Subarasheese
  14

• DDP GPU0 allocate too many memory

  #36 opened a year ago by yijingshihenxiule
  8

• Inference error, scipy

  #37 opened a year ago by AWAS666
  2

• Are the parameters for the Duration Discriminator being incorrectly passed?

  #30 opened a year ago by innnky
  8

• Failed to do inferencing with the provided checkpoint and inferencing code

  #24 opened a year ago by Subarasheese
  4

• log_w_ not detach

  #21 opened a year ago by hdmjdp
  2

• custom inference with vits2

  #13 opened a year ago by icklerly1
  13

• why not net_dur_disc with DDP in train_ms.py?

  #23 opened a year ago by hildazzz
  1

• 44100Hz Ja Support

  #19 opened a year ago by tonnetonne814
  0

• This is not an issue but results of using vits2_pytorch

  #17 opened a year ago by athenasaurav
  3

• a trivial bug

  #16 opened a year ago by 16dian11
  1

• Does the speaker embedding conditioning work?

  #15 opened a year ago by BangjianZhou
  1

• Error in code when training

  #14 opened a year ago by Silver267
  2