p0p pflowtts_pytorch Issues

p0p4k/pflowtts_pytorch Issues

High duration loss
Updated 3 months ago15
about zero-shot inference
Updated 4 months ago31
Hi, Can I make longer pauses between sentences? Is it possible or not?
Updated 4 months ago2
phoneme question
Updated 4 months ago7
pflow performance
Closed 4 months ago
Is there any already trained english (or any other) model, that we can test, maybe based on ljspeech dataset?
Updated 4 months ago1
Crash in MAS
Updated 4 months ago22
About model implementation differences
Updated 4 months ago3
2nd-order Heun's method for ODE sampling
Updated 4 months ago2
NAN loss
Updated 4 months ago8
about n_spks
Updated 4 months ago6
Error while finetuning multi speaker from pretrained model
Updated 5 months ago3
How to perform fine-tuning?
Closed 5 months ago7
Does model output phoneme-level timing info ?
Updated 5 months ago1
Works only with n_feats=80
Updated 5 months ago6
noise scaled mas
Closed 5 months ago4
Error when training single speaker model from scratch
Closed 5 months ago2
Error when finetuning multi speaker from pretrained model
Closed 5 months ago1
Unused model
Closed 6 months ago3
Loss masking?
Closed 6 months ago15
Multi gpu training; RuntimeError: [...] LightningModule has parameters that were not used in producing the loss returned by training_step.
Updated 6 months ago3
AlignerNet
Closed 6 months ago8
Distortion of audio
Updated 6 months ago2
Noise in e2e branch
Closed 7 months ago2
Prior or encoder Loss
Closed 7 months ago2
Positional encoding
Updated 7 months ago1
descript/encodec is too slow in dataloader
Updated 7 months ago7
Jump in sub_loss/train_dur_loss_step
Updated 7 months ago6
it's possible voice change to clone new voice with just one wav file or more ?
Closed 7 months ago4
Make it multi-language?
Updated 7 months ago10
_clean_text is returning invalid symbol
Updated 8 months ago6
The pre-trained hifigan model
Closed 8 months ago3
pflow/data/text_mel_datamodule.py __getitem__
Closed 8 months ago7
TypeError: pflowTTS.synthesise() got an unexpected keyword argument 'spks'
Updated 8 months ago2
possible custom model inference issue
Updated 8 months ago1
How ready is multispeaker training?
Closed 8 months ago1
requirements & phonemizer ?
Closed 8 months ago2

p0p4k / pflowtts_pytorch

p0p4k/pflowtts_pytorch Issues

High duration loss

about zero-shot inference

Hi, Can I make longer pauses between sentences? Is it possible or not?

phoneme question

pflow performance

Is there any already trained english (or any other) model, that we can test, maybe based on ljspeech dataset?

Crash in MAS

About model implementation differences

2nd-order Heun's method for ODE sampling

NAN loss

about n_spks

Error while finetuning multi speaker from pretrained model

How to perform fine-tuning?

Does model output phoneme-level timing info ?

Works only with n_feats=80

noise scaled mas

Error when training single speaker model from scratch

Error when finetuning multi speaker from pretrained model

Unused model

Loss masking?

Multi gpu training; RuntimeError: [...] LightningModule has parameters that were not used in producing the loss returned by training_step.

AlignerNet

Distortion of audio

Noise in e2e branch

Prior or encoder Loss

Positional encoding

descript/encodec is too slow in dataloader

Jump in sub_loss/train_dur_loss_step

it's possible voice change to clone new voice with just one wav file or more ?

Make it multi-language?

_clean_text is returning invalid symbol

The pre-trained hifigan model

pflow/data/text_mel_datamodule.py getitem

TypeError: pflowTTS.synthesise() got an unexpected keyword argument 'spks'

possible custom model inference issue

How ready is multispeaker training?

requirements & phonemizer ?