新版本是在Illustrious模型上训练的，我没有测试它与其他模型的兼容性，不能保证它能正常工作。

The new version was trained on the Illustrious model, I didn't test it for compatibility with other models and can't guarantee it will work.

这个Lora并不包括动画中全部的角色，但你可以尝试通过相似的Tag来生成相似的角色。这个Lora风格不稳定，不建议新手使用，如果你比较熟悉Tag的使用，你可以生成非常有趣和动感的效果。

This Lora does not include all the characters in the animation, but you can try to generate similar characters by using similar tags. This Lora style is unstable and not recommended for beginners, if you are more familiar with the use of Tag, you can generate very interesting and dynamic effects.

不要设置太高权重，否则会模糊。

Don't set the weight too high or it will blur.

我推荐0.6~1.0之间。

I recommend that between 0.6 ~ 1.0.

If you want characters, you can try：panty & stocking with garterbelt