For the SDXL version click here.
Want to commission a LoRA? Hit up my ko-fi. I'll find the images and train the model.
Nearly every image in the galleries were done with Realistic Vision 2.0.
I need to remake with better captioning and more epochs. Undercooked.
Description
FAQ
Comments (6)
what model did u use to generate these images?, they look great btw
Mostly all Realistic Vision 2.0. A checkpoint designed for anime or 2d might look better. And thank you.
What models are being used in these? Im using the realistic vision 2.0 model but cant get it like the preview pictures.
https://i.imgur.com/Lmphsjz.png This is one result i got.
Looks like you don't have the <lora:nori:1.0> token.
And I may have used a weight of 1.2 or 1.4, since I think the LoRA was undercooked.
@chairfull That was it sorry. Got them working now but ill try it higher. thank you!