PyTorch Implementation of Meta-StyleSpeech : Multi-Speaker Adaptive Text-to-Speech Generation
fix optimizer (Disc) and update pre-trained model
fix duration control at inference time
First Official Release