AI净土交流群电报：<a target="_blank" rel="ugc" href="https://t.me/+soU-Ws67aBlkY2I1">https://t.me/+soU-Ws67aBlkY2I1</a>凯登AI私模定制咨询电报：<a target="_blank" rel="ugc" href="https://t.me/+Ql4sngcJlbhmOTU1">https://t.me/+Ql4sngcJlbhmOTU1</a>Trigger Words : dunhuang 触发词 : dunhuang<h2 id="heading-1264">V3版说明：</h2>1.对服饰的一致性进行修改，整体画风不再杂乱2.构图不容易崩坏手脚3.服饰更加自然4.对动画模型和真实模型的兼容性更好总体而言更容易出好看的作品<h2 id="heading-1265">Version V3 description:</h2>1. Modify the consistency of clothing, and the overall painting style is no longer tangled2. Composition is not easy to break hands and feet3. Dress more naturally4. Better compatibility between animated models and real modelsIt's easier to produce beautiful work in general加入上百种敦煌风格服饰，降低脸部权重，更好与lora搭配！<h3 id="heading-1266">服装分层数据：</h3><h3 id="heading-1267">CLOTHING:1,1,1,1,1,0,0.2,0,0.8,1,1,0.2,0,0,0,0,0</h3>注意：示例图片中并未对人物手以及姿态和表情进行调节，展示图片也并未精心挑选，而是随意生成并展示几张服饰效果This version adds over a hundred Dunhuang style clothing styles to reduce facial weight and better match Lora! Note: The example images did not adjust the hands, posture, and expressions of the characters, and the display images were not carefully selected. Instead, several clothing effects were randomly generated and displayed <h1 id="heading-1268"><s>V2.1版本</s></h1>一些示例：（由于舞蹈动作手部肢体特征较多，经常崩坏，案例图片我也懒得修复了）去掉背景只保留人物服饰特征：&lt;lora:dunhuangV2.1:1:1,1,1,1,0,0,1,1,1,1,1,1,1,0,0,1,1&gt; 以下是去除敦煌风格背景，再重绘背景的效果<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/9dfbff2c-691b-45fe-87c8-cd2614620134/width=525/9dfbff2c-691b-45fe-87c8-cd2614620134.jpeg" /><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/8c25e74a-ed72-4c99-9754-a215de54cd29/width=525/8c25e74a-ed72-4c99-9754-a215de54cd29.jpeg" />

Files with SHA256: 702841a6a5d128a5ca3894f44f50490fb495fb46fe29b09a92b1bad6061b77dc