用戶LoRA
하나나기
semi-realistic anime style, clean lineart, soft rendering, smooth skin, subtle gradients, transparent shading, soft lighting, gentle shadows, crisp facial details, beautiful eyes, novel illustration
This LoRA is trained to reproduce the illustration style shown in the preview images.
This LoRA is intended for style transfer only and can be combined with your own character prompts. Recommended Settings
Base Model: Light Model Recommended Weight: 1.2 CFG: 5–7 Steps: 25–35 Sampler: DPM++ 2M Karras
