You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am lloking for better configurations for training a Hindi language TTS.
I have used Ossian with 16 layers of NN each for accoustic and Duration training with 25 and 100 epochs respectively using naive_01_nn recipe, but there is no significant change to the previous configurations.
So if you could please post a good configuration or a new recipe.
The text was updated successfully, but these errors were encountered:
I am lloking for better configurations for training a Hindi language TTS.
I have used Ossian with 16 layers of NN each for accoustic and Duration training with 25 and 100 epochs respectively using naive_01_nn recipe, but there is no significant change to the previous configurations.
So if you could please post a good configuration or a new recipe.
The text was updated successfully, but these errors were encountered: