Based on 96 images dataset, created lora for 4 different model with varying results
Keyword: sc4r.
for zimage / flux anime keyword help: digital drawing, describing a feral lion (to evoid having a human or too realistic plush). best model are still anima/illustrious