LORA for semi-realistic, trained based on Illustrious0.1
strengsth 0.5~1。reccomend 0.7
il_one.safetensors