! Experimental LoRA !
I tried to make Carnelian's newer style LoRA because her baked style in Illustrious is from her older one. It's not perfect LoRA as my dataset is pretty limited so expect some artifacts. And I accidentally used Illustrious 1.0 as base model lol.
Best use on Illustrious-based model/checkpoint, use strength at 0.7-1.