我想做一个让场景概念笔触更加概括的模型，拉开场景虚实关系。于是用一些场景概念图训练了这个Lora模型。在用一些绘画类模型时有一点效果，感兴趣的可以自己试试。

I want to make a model that allows the scene concept strokes to be more generalized, pulling apart the scene virtual and real relationships. So I trained this Lora model with some scene concept drawings. It has a little effect when using some painting class models, so you can try it yourself if you are interested.（Translated from DeepL）

关键词Keywords：（concept）（sutu）