NVIDIA/waveglow

Issues

• Citation of waveglow_256channels_v5

  #226 opened 4 years ago
  4

• Does waveglow model the spectrum magnitude only?

  #225 opened 4 years ago
  2

• device=cpu, still tries to use GPU?

  #224 opened 4 years ago
  1

• NaN or Inf found in input tensor

  #223 opened 4 years ago
  5

• Error while loading a waveglow checkpoint for inference

  #222 opened 4 years ago
  3

• [Question] published latency/delay of WaveGlow vocoder for real-time streaming inference

  #220 opened 4 years ago
  1

• Waveglow on multi speaker

  #219 opened 4 years ago
  2

• explanation of hparams and config.json

  #218 opened 4 years ago
  1

• Training loss is Not decreasing and saturates around ~-5, -6 when training with Librispeech dataset(16KHz)

  #217 opened 4 years ago
  1

• Input of the Waveglow

  #216 opened 4 years ago
  0

• Training waveglow model for 16kHz

  #215 opened 4 years ago
  15

• use .npy file to get the audio

  #214 opened 4 years ago
  6

• About versions of published models

  #213 opened 4 years ago
  1

• Why does the Nvidia results page show contradictory results for inference?

  #212 opened 4 years ago
  11

• How many flops in the inference for 1 second of audio?

  #211 opened 4 years ago
  5

• Is there a way to increase speaking speed via WaveGlow Vocoder?

  #210 opened 4 years ago
  3

• file errors in windows

  #209 opened 4 years ago
  1

• not works

  #208 opened 4 years ago
  2

• requirements? (python,os)

  #207 opened 4 years ago
  1

• Reducing n_mel_channels to 40: Error of size mismatch

  #206 opened 4 years ago
  2

• Gradient overflow when training

  #205 opened 4 years ago
  18

• waveglow_256channels.pt

  #204 opened 4 years ago
  6

• pretrain model

  #203 opened 4 years ago
  3

• Model can approximately fit but has consistent artifact on every audio synthesis

  #202 opened 4 years ago
  4

• Excess padding

  #201 opened 4 years ago
  2

• Warm start from published model

  #200 opened 4 years ago
  15

• ImportError: cannot import name init_distributed

  #199 opened 4 years ago
  1

• How much the loss should reduce before reasonable sound can be produced?

  #198 opened 4 years ago
  5

• Training with raw un-normalized mel features was resulting in a clean synthesis

  #197 opened 4 years ago
  2

• Need some explainations in hparams.py

  #196 opened 4 years ago
  0

• Text2Mel input to WaveGlow outputs noisy audio file without any speech

  #195 opened 5 years ago
  1

• Optimize model for inference speed

  #194 opened 4 years ago
  6

• Aligning conditioning information and squeezed input waveform

  #193 opened 5 years ago
  2

• Inference Error：ValueError: invalid buffering size

  #192 opened 5 years ago
  1

• Can a trained model continue training with different number of GPUs?

  #191 opened 4 years ago
  4

• Cant load waveglow checkpoint into inference code even after convert

  #190 opened 5 years ago
  37

• How to train new Waveglow model for diffirent language?

  #189 opened 4 years ago
  17

• How to load local model using torch.load instead of torch.hub.load?

  #188 opened 5 years ago
  3

• RuntimeError: Given groups=1, weight of size 256 3 1, expected input[1, 2, 25472] to have 3 channels, but got 2 channels instead

  #187 opened 5 years ago
  1

• Distributed training no faster... just gets more conservative

  #186 opened 5 years ago
  3

• CUDA out of memory，even set batch_size to 1

  #185 opened 5 years ago
  8

• Trying to train on LJSpeech data, it does nothing.

  #184 opened 5 years ago
  3

• Scaling learning rate for multigpu training

  #183 opened 5 years ago
  0

• Deploy without GPU?

  #182 opened 5 years ago
  3

• Here is my hypothesis while training dataset.

  #181 opened 5 years ago
  4

• How to generate speech directly from mel spectrograms of custom audio files

  #180 opened 5 years ago
  8

• Training on LJ Speech data gets stuck

  #179 opened 5 years ago
  0

• How do I output data to tensorboard?

  #178 opened 5 years ago
  2

• python synthesis.py ${checkpoint_path} ${text_list.txt} ${output_dir} --preset=<json>

  #177 opened 4 years ago
  1

• Generate a package tts

  #176 opened 5 years ago
  1