Description
This model is a Finetune of v-pred 1.0 aiming to stabilize it, specially guiding it away from extremely dark or extremely bright images for no aparent reason, quite flexible with booba thicc and futa data in it.
Blame @justTNP for this model's name.
Recipe for this model was: Finetuned then mixed into itself with alpha 0 to tweak brightness and contrast a bit.
This is a v-Pred zSNR model, but has the empty tensors for detection on reforge and other UIs, please only use v-pred LoRAs on this.