A simple style LoRA made from a dataset of generated images from my 1.5 models. Works well at weights 0.8 - 1.0 I recommend 0.8