/vt/ - Virtual Youtubers

>>62670129
Dreambooth and Finetune are different methods afaik. Dreambooth uses regularization images or the class to distinguish between stuff but it works with less images in the dataset. I think finetuning might be closer to hypernetworks. I don't know for sure, just testing stuff out. They're separate methods on kohya's anyway.
I didn't get the chance to try out pivotal tuning yet because it requires a different trainer and had too much stuff to do to set it up but maybe tomorrow.