Fahrettin dataset

[FatihcanUslu]

Hello,

I am currently using the Fahrettin Turkish model from Piper, and it performs very well overall. However, when I fine-tune the model using my own voice, I notice an unusual background noise or crackling artifact in the generated audio.

I am wondering whether this issue could be related to the training data quality or preprocessing steps on my end. In this context, I wanted to ask is the Fahrettin dataset available as open source?