Quality of samples doesn't match demos

[rezareza007]

Hi,

I tried to convert some sentences to speech using this fork: https://github.com/Mehrdadghassabi/Persian-text-to-speech However the quality of synthesized audio was not anywhere close to the samples on Soundcloud. I don't know if I'm doing anything wrong.

Also I wanted to see how I can use another vocoder, such as WaveGlow, instead of Griffin-Lim?

[Mehrdadghassabi]

Hello Reza I hope you are doing well, looks like this model doesnt have enough qualification, maybe due to lack of training data or poor design