This checkpoint was created by merging my custom-trained Doc's Dark Fantasy LoRA into Stable Diffusion XL Base 1.0 at a strength of 0.8. The original LoRA was trained on 77 reference images over 2,000 steps.