Search results

Training StyleTTS2 on Indic Languages or Code-Mixed Data – Any Success Stories?

Status: Open.

#334 In yl4579/StyleTTS2;

· Ramakrishna-Chaitanya opened on Jun 6, 2025

Train second stage on single GPU with LJSpeech dataset, but stop with g_loss NaN

Status: Open.

#333 In yl4579/StyleTTS2;

· renwotao opened on Jun 3, 2025

Train first stage on LJSpeech， train gen_loss is slower increase？？？Is this correct？？

Status: Open.

#332 In yl4579/StyleTTS2;

· renwotao opened on May 23, 2025

针对于中文参考音频以及中文参考文本

Status: Open.

#331 In yl4579/StyleTTS2;

· lhhssd opened on May 19, 2025

Speaking rate change in consecutive trainings

Status: Open.

#329 In yl4579/StyleTTS2;

· Ananya21162 opened on Apr 22, 2025

Would you mind offer the 'Models/LJSpeech/first_stage.pth' checkpoint for first stage?

Status: Open.

#328 In yl4579/StyleTTS2;

· CindyTing opened on Apr 18, 2025

Getting NaN in Mel Loss during the first few epochs for first training

Status: Open.

#327 In yl4579/StyleTTS2;

· iAdityaVishnu opened on Apr 17, 2025

Request for Guidance on Training a Turkish Voice Model with StyleTTS2

Status: Open.

#325 In yl4579/StyleTTS2;

· ooaykac opened on Apr 8, 2025

JDCNet model input shape and features

Status: Open.

#324 In yl4579/StyleTTS2;

· Mayank-Gupta11 opened on Apr 1, 2025

GPU memory error occurs at epoch 50 during phase first stage model training.

Status: Open.

#321 In yl4579/StyleTTS2;

· kadirnar opened on Mar 9, 2025

Consistently getting CUDNN_STATUS_INTERNAL_ERROR during second_stage training after diff_epoch

Status: Open.

#319 In yl4579/StyleTTS2;

· hanshounsu opened on Mar 8, 2025

First step training very slow and high GPU memory

Status: Open.

#318 In yl4579/StyleTTS2;

· zaidato opened on Mar 7, 2025