The result of different "outputs_per_step" in hparam.py

keithito / tacotron

A TensorFlow implementation of Google's Tacotron speech synthesis with pre-trained model (unofficial)

MIT License

2.95k stars 959 forks source link

Closed AppalachianWine closed 4 years ago

AppalachianWine commented 4 years ago

Hello, when I use train the tacotron model, the effect of using r = 2 is better than r = 5. Does anyone have any conclusions about this?