The model I think is fairly self explanatory.
Works fine at a strength of 1. Previews made with novaFurryXL_illustriousV20.
Description
Attempt to generate their non-ai style.
FAQ
Comments (9)
I don't know if you've been asked this before, but why don't you use NoobAI as the base model for the loras you train? Or do you consider the model you're using to be better?
The base model I use is illustrious the default model for baking, I'm not sure if it was my baking settings before I experimented and figured out what actually worked or if it was my prompt, but if you look back at say for example the gwen model I did, there's all sorts of errors with the images that don't seem to really be solvable from the mild testing I did when I baked with noobai epsionpred1 and 1.1 and then generated images using those models, I generate preview images in the gallery with NovaFurryXL because it seems to be a relatively neutral model that adapts to styles fairly well like how autismmix did for confetti. Currently what I do seems to work a lot better than anything I did in the past, but I'll try fiddling around with some of the noobai recent versions with some small quick loras to see if it was my baking settings or if it's the prompt or checkpoint that was the issue now that I've got a great baking preset
my goat the trap doujinshi models uve been cracking recently have been gas, do u have a request form or sth somewhere cause theres definetly some missing that id LOVE to see
Just send me a direct message, it's more reliable than comments. the best way to request things is usually to send me a direct message. There's no formal request forms or anything. I typically do requests as they come from people just asking.
@Malebolgia sure thing !
Can you train shexyo before he started using ai? That version personally is very different to what he does now.
Should be ~2 hours.
It's done!
@Malebolgia SO MUCH better.Making post rn