Rayhane-mamah / Tacotron-2

DeepMind's Tacotron-2 Tensorflow implementation

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

why the los of the wavenet training is changing around -2.54 to -4.08 it is not going further and even there is no improvements in the results of wavenet.

kingsyard opened this issue · comments

as it is becomming hard i cant able to get the results imroved even after 170k steps of wavenet training.why loss is very very unstable .