enhuiz/vall-e
An unofficial PyTorch implementation of the audio LM VALL-E
Leverages Meta's EnCodec tokenizer to compress audio into discrete codes, then trains separate autoregressive and non-autoregressive transformer models for zero-shot text-to-speech synthesis from reference audio. Training uses DeepSpeed for distributed optimization across quantization levels, with a complete pipeline for data preprocessing (audio quantization, phoneme generation), model export, and synthesis inference.
2,992 stars. No commits in the last 6 months.
Stars
2,992
Forks
406
Language
Python
License
MIT
Category
Last pushed
May 10, 2023
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/voice-ai/enhuiz/vall-e"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
bshall/Tacotron
A PyTorch implementation of Location-Relative Attention Mechanisms For Robust Long-Form Speech Synthesis
DemisEom/SpecAugment
A Implementation of SpecAugment with Tensorflow & Pytorch, introduced by Google Brain
Kyubyong/tacotron
A TensorFlow Implementation of Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model
Rayhane-mamah/Tacotron-2
DeepMind's Tacotron-2 Tensorflow implementation
Kyubyong/dc_tts
A TensorFlow Implementation of DC-TTS: yet another text-to-speech model