I tested it using the official FP8 model. It is recommended to set the LoRA strength to 0.8. When the strength is set to 1 and combined with other NSFW models, it will produce a large amount of NSFW content.
Flat style_v1.safetensors