Skip to content

Latest commit

 

History

History
6 lines (4 loc) · 215 Bytes

README.md

File metadata and controls

6 lines (4 loc) · 215 Bytes

tacotron-2(pytorch) + wavernn(pytorch) chinese TTS:

chinese TTS , tacotron2 support n_frames_per_step , convert wavernn model to c++ inference.

location sensitive attention
stepwise monotonic attention