GLaDOS (Portal, 200 epochs, rmvpe)

[rough urchin]

Link: https://huggingface.co/Homiebear/GLaDOSRetrained/resolve/main/GLaDOSV2.zip?download=true

Voice Lines Ripped Directly from:
-Portal 1
-Portal 2

Dataset: 20:01
Sample Rate: 32k
Pretrain: Original
Recommended Pitch: 0
Batch size: 4
Made on Kaggle

---

Credit:
If you use this model please credit me on youtube: @Th4t-Ai-Guy

[magic gull]

I need the original

[raw ermine]

why do u guys add reverb on every sample?

[magic gull]

I don't, I just use Astra labs

[rough urchin]

I used astra lab for 1 sample..

1 SAMPLE HAS REVERB YOU KNOB.. Mb mb

[raw ermine]

remov dem reverb ;-;

[magic gull]

literally dunno how

[raw ermine]

use ilaria or applio to test dem models

[magic gull]

too lazy to remove the music to get clean samples

[raw ermine]

bruh

[magic gull]

I'd use Weights but uploading here auto uploads to weights

[raw ermine]

don't forget to extract the back/lead vocals tho if ur using weights o - o