重要提示：

1. 请通过负数UNet权重加载！

或将该LoRA置于负面提示词中。

Please use it with negative UNet weights!

Or use this LoRA in the negative prompt.

2. 请在noobaiXLNAIXL及其子模型中使用，我无法保证其在其他模型中的效果！

Please use it with noobaiXLNAIXL and its sub-models. I cannot guarantee the effectiveness in other models!

为何不用embedding？只是因为我个人不喜欢，更倾向于通过LoRA调用的形式。

Why not embedding? Just because I don't like it, I tend to prefer the form of calling through LoRA.

推荐权重：(-0.3)-(-1.0)

Recommended weight: (-0.3)-(-1.0)

模型效果参考以下示例：

Model effect reference with the following example:

---