NATSpeech Versions Save

A Non-Autoregressive Text-to-Speech (NAR-TTS) framework, including official PyTorch implementation of PortaSpeech (NeurIPS 2021) and DiffSpeech (AAAI 2022)

pretrained_models

2 years ago

This release includes LJSpeech pre-trained models:

  • hifi_lj.zip: HiFi-GAN
  • ps_normal_exp.zip: PortaSpeech (normal version)
  • ps_small_exp.zip: PortaSpeech (small version)
  • ds_exp.zip: DiffSpeech
  • aux_exp.zip: a pre-trained FastSpeech2 checkpoint for training DiffSpeech