This LoRA is trained to imitate the style of Mirco Cabbia, aka Sciamano240.

For the best results, use the model "Realistic Vision V1.3", check Hires-fix and set 0.2 for denoising strength, and choose SwinIR_4X as upscaler.