Description

- Harvest pitch extraction (I made this before being aware of this server, might retrain with even more data and crepe-mangio later on) - 200 epochs, trained on a total of 20 minutes of her speaking, I prepared the dataset hand picking the best sounding parts of streams and treating the audio to the best of my ability while keeping clarity high. - Model and data set made by me, will most likely keep updating it until I'm fully satisfied with it. If you do use it, credit would be appreciated. ^^