Description

(70 epochs was when the "loss/g/total" graph reached the minimum. I may test the model at other epochs.) The model was trained by me using Mangio-Crepe-RVC-Fork and mangio-crepe for feature extraction. About 12 minutes of voice clips from *Team Sonic Racing* and *Mario & Sonic at the Olympic Games Tokyo 2020* were used. Credit if reused. Google Drive Face link NEW! A version of the model trained by Kits AI: