Link - https://huggingface.co/Freaky98/KinitoPET_rvc-v2-120e.zip/resolve/main/KinitoPET_rvc-v2-120e.zip?download=true

This is a definitive version of KinitoPET. The model was made from 40 minutes (33 without silence) of a lossless dataset that was meticulously manually cleaned from all sorts of buzzing and noises. The sample rate is 48k.
There won't be a better version of this voice.
The model also works great with real-time voice changer.

I recommend setting the feature ratio at 1.

Additional Details:

Tags: No tags available

Download Link: https://huggingface.co/Freaky98/KinitoPET_rvc-v2-120e.zip/resolve/main/KinitoPET_rvc-v2-120e.zip?download=true