GuideVocalPretrain (RVC v2) | AI HUB | Page 1

maiden elm Dec 15, 2024, 12:53 AM

#

batch_size : 16
Epoch & Steps: 22 Epochs, 102498 Steps
RVC version: v2
Dataset Length : 58 hrs (Korean Guide Vocal Dataset - Normal)
F0 pitch Extraction Method: RMVPE
Sample Rate: 48khz only

Finetuned from Original Pretrain.

DL Link:

G: https://huggingface.co/Essid/GuideVocalPretrain/resolve/main/G_GuideVocalPretrain.pth
D: https://huggingface.co/Essid/GuideVocalPretrain/resolve/main/D_GuideVocalPretrain.pth

#

Trained Model Audio Sample

gritty falcon Dec 15, 2024, 10:34 PM

#

what does it do?

fervent lion Dec 19, 2024, 7:08 AM

#

another korean pretrain o - o

hazy wyvern Dec 19, 2024, 7:16 AM

#

maiden elm Trained Model Audio Sample

tetoris

#GuideVocalPretrain (RVC v2)