This LoRA model is primarily used to restore the luminous aspects of light novel characters, you can add prompt such as black dress, silver long hair, fangs, and black hair accessories to enhance the visual effect.
The optimal weight range is around 0.5 to 0.8. Weights below 0.6 can more accurately depict action scenes, while weights of 0.8 can more accurately reproduce facial details, pupils, and clothing. The best results are achieved by combining with image to image .