<h3>基础信息</h3><p>四肢欠损光辉模型，保留一定残肢，支持一定幅度的动作，触发词为:</p><pre><code>quadamp</code></pre><h3>训练参数</h3><p>训练模型：<strong>WAI-illustrious-SDXL-v16.0</strong></p><p>轮数：12</p><p>最终损失值：0.0741</p><p>训练集大小：77</p><p>推荐分辨率：768x1344，896x1152</p><h3>试验性内容</h3><p>基于数据集大小限制，在参数提示词中加入了关于腿部残肢站立（standing on stumps）与手部残肢抓拿物品（stump holding），但效果欠佳，在高权重情况下会使画面中出现树桩（stump），请谨慎使用。最佳使用提示词：</p><pre><code>full body,sitting,short sleeves</code></pre><p>等搭配使用得到最佳效果。</p><h1 id="basic-information"><strong>Basic Information</strong></h1><p>Quadruple amputee radiant model, retains certain residual limbs, supports a certain range of motion. Trigger word is:</p><pre><code>quadamp</code></pre><h1 id="training-parameters"><strong>Training Parameters</strong></h1><p>Training Model: <strong>WAI-illustrious-SDXL-v16.0</strong><br />Epochs: 12<br />Final Loss Value: 0.0741<br />Dataset Size: 77<br />Recommended Resolutions: 768x1344, 896x1152</p><h1 id="experimental-content"><strong>Experimental Content</strong></h1><p>Due to the limited size of the dataset, prompts involving standing on leg stumps ("standing on stumps") and hand stump grasping ("stump holding") were included in the training captions. However, the results are suboptimal; at high prompt weights, this may cause actual tree stumps to appear in the image. Please use with caution. Optimal usage prompts include:</p><pre><code>full body, sitting, short sleeves</code></pre><p>Combine these for the best results.</p>

Files with SHA256: e1f8e37c3d74f6a3429319205e9ea3ffcf8df52fbdf121d0ba610a5f1095216b