ParallelWaveGAN

ParallelWaveGAN copied to clipboard

Low inference speed of TTS on GPU

2

May I ask why the RTF of TTS is only 0.09 for a 12-seconds sentence? I use fastspeech2_HIFiGAN model and GPU is A2000 (8.0 capability). I thought it should be...

dalvlv

A new interesting vocoder was described in a paper yesterday. It's called Avocodo and supposedly helps with the artifacts that are typical for GAN based vocoding. It supposedly also works...

Flux9665

stylemelagn model structure change for 32Khz vocoder

3

Dear professor, I'd like to train stylemelgan vocoder of 32kHz, here is my config to train a multi-speaker model, now the speaker similarity on VC task is worse than fregan/hifigan....

wblgers

Trills and jitters in wav synthed by mb-melgan

1

@kan-bayashi I train mb-melgan with baker or LJ speech dataset, but synthed wavs seem to have distinct trills and jitters. Any ideas or advices to solve this probem? thanks

xiaoyangnihao

I'm fine tuning Hifigan from 2.5ml steps pretrained model to 3ml steps. I wonder if this is the way to go by updating milestones? ``` generator_optimizer_type: Adam generator_optimizer_params: lr: 2.0e-4...

skol101

HifiGAN training -- obvious harmonics in test files

7

I trained HifiGAN on VCTK multi-speaker datasets with 24kHz sampling rate. I also do normalization in the input log-Mel spectrogram (with mean=-4, std=4)， and found obvious harmonics in the test...

Kristopher-Chen

How to solve the problems?

2

Dear professor: I have already install the ParallelWaveGAN,however,when I run the egs/vctk/voc1 run.sh, it comes with the error: Stage 0: Data preparation Successfully split data directory. ERROR: num_first + num_second...

Hu-chengyang

In stylemelgan, G net and D net joint training, voice quality decreased

4

Hi，sir,I have a problem,and Could you help me?I choose the generative network of stylemelgan and the discrimination of melgan. When I pre train the generated network, the voice quality is...

huhuqwaszxedc

Benchmark RTF for all method: Melgan - MBMelgan - Hifigan - StyleGAN

1

I have 3 questions to confirm my experiment on these vocoder models: 1: Are there anyone have a comparing summaries between all model in RTF (with light and heavy version)...

v-nhandt21

Breakpoint problem

22

hi @kan-bayashi , When using this project to train multi-band melgan vocoder, there will always be breakpoints in the generate audio! Trying to modify the kernal size of the first...

maozhiqiang