<ul><li>该模型基于 Animagine XL V3.0 Base 底模训练</li><li>用法：<ul><li>无需提示词触发。</li><li>推荐大模型：Animagine XL V3.1。</li><li>不使用 VAE。</li><li>采样器：Euler a，采样步数 28 步。</li><li>CFG Scale：6-7。</li><li>T2I 分辨率：不低于 768。</li><li>Lora 权重：1.0。</li><li>开启 ADetailer 可以更好地修复手和脸。</li></ul>注意事项：<ul><li>衣服在训练时作为训练内容，可能难以替换。</li><li>如果出现 NSFW 结果，请在正面提示词中添加 "jacket" 或更多文本/场景补充。</li><li>如果头发出现奇怪的颜色，请在负面提示词中添加 "multicolored hair"。</li></ul>该 Lora 模型的训练数据集截取自番剧《GIRLS BAND CRY》第一集，采用标准优化器 AdamW 训练。欢迎使用其他模型（例如：animelllustDiffusion_v08）并分享你的成果。</li></ul><ul><li>This model is based on the Animagine XL V3.0 Base model.</li><li>Usage:<ul><li>No prompt required.</li><li>Recommended base model: Animagine XL V3.1.</li><li>No VAE.</li><li>Sampler: Euler a, 28 sampling steps.</li><li>CFG Scale: 6-7.</li><li>T2I resolution: Not less than 768.</li><li>Lora weight: 1.0.</li><li>Enabling ADetailer can improve the generation of hands and faces.</li></ul></li><li>Notes:<ul><li>Clothes may be difficult to replace as they were part of the training data.</li><li>If NSFW results are generated, add "jacket" or more text/scene details to the positive prompt.</li><li>If hair has strange colors, add "multicolored hair" to the negative prompt.</li></ul>This Lora model was trained on the first episode of the anime "GIRLS BAND CRY" using the standard optimizer AdamW. Feel free to use other models (e.g., animelllustDiffusion_v08) and share your results.</li></ul>

Files with SHA256: 64a58a0b59ca994e491767c1d2fb8d2f56d7e89d3cfd844710d42ccd906e7f11