tts_impl

pytorch implementation of speech synthesis methods

Work in progress

Models

implemented: ✅
work in progress: 🚧

• Vocoders
  • HiFi-GAN ✅
  • HnNSF-HiFi-GAN ✅
  • ISTFTNet
  • WaveNeXt 🚧
  • Vocos 🚧
  • DDSP (Subtractive / Additive)
• TTS: Text to speech
  • VITS ✅
  • VITS2 🚧
  • JETS 🚧
  • FastSpeech
  • Diffusion TTS Models
  • Flow Matching TTS Models
• VC: Voice Conversion
  • VITS-SVC
• SVS: Speech Voice Synthesis
• PE: Pitch Estimation
• Aligner:
  • ForcedAlign ✅

Installation

pip install git+https://github.com/uthree/tts_impl.git

or clone this repository,

pip install -e .

code check

black src; isort src to run formatter
pytest to run test
mypy src to run type check