About the performance of synthesis

Kyubyong / tacotron

A TensorFlow Implementation of Tacotron: A Fully End-to-End Text-To-Speech Synthesis Model

Apache License 2.0

1.83k stars 436 forks source link

Closed CherryCloris closed 6 years ago

CherryCloris commented 6 years ago

I used the model synthesis, but it sounds not very good, it feels like an echo. How about you?