Open oyeamit opened 4 years ago
this code use a teacher-student mechanism. the autoregressive model just used as a teacher, to generate phoneme durations to train the student forward model. so you maight to modify the coressponding code about the duration extraction
Hi, yes you will want to train a forward model for this. There you can easily directly control the duration of each phoneme
First of all, Great Work! Thanks for sharing the repo!
I have trained the autoregressive model on LJ dataset. The output is quite good for short sentences. I seek some advice to manipulate pauses between words in the produced speech. Let's say the produced speech is 'This is Text to Speech model.' I want to increase(or say decrease) the pause between the word Speech model little bit.
Any Suggestions?