LoRA model of 6th Anniversary Costume Future Signature (フューチャー・シグネチャー) of mobile game THE IDOLM@STER MILLION LIVE! THEATER DAYS (アイドルマスターミリオンライブ！シアターデイズ).The dataset contains unofficial materials, and I specially train the hair bow/papillon emblem/pocket/back ribbon/shoes, which may help ease the overfitting and improve the detiails of picture.The trigger word is future signature (costume), with three colorways available:*Princess style:<pre><code>future signature (costume), pink clothes, pink hair bow, pink footwear</code></pre>*Fairy style:<pre><code>future signature (costume), blue clothes, blue hair bow, blue footwear</code></pre>*Angel style:<pre><code>future signature (costume), yellow clothes, yellow hair bow, yellow footwear</code></pre>You can replace the color of tags above, and it will work.The LoRA weight of samples is 0.8-0.85; you can also try other weights.Note: I load my LoRA model through the additional network WebUI plugin, so the prompt of sample illustrations don't contain &lt;lora:XXX:X&gt; tag. Color pollution and decoration missing may occur when there are multiple subjects in the illustration.I'll appreciate it if you're willing to use this LoRA model and share AI arts to other people. My models and illustrations are free for everyone, and if you like them, please support me at <a target="_blank" rel="ugc" href="https://www.patreon.com/Kaorin_plkm160">Patreon</a> or <a target="_blank" rel="ugc" href="https://afdian.net/a/KaorinP_plkm160">爱发电</a>.————————————手机游戏《偶像大师 百万现场 剧场时光》（MLTD）六周年服装フューチャー・シグネチャー（Future Signature）LoRA模型模型的数据集包含非官方素材，而且我专门训练了蝴蝶结发饰/蝴蝶徽章/口袋/背带/鞋子的样式，或许可以帮助缓解过拟合并提高画面的细节。模型的触发词为 future signature (costume)，带有三种配色：*Princess样式：<pre><code>future signature (costume), pink clothes, pink hair bow, pink footwear</code></pre>*Fairy样式:<pre><code>future signature (costume), blue clothes, blue hair bow, blue footwear</code></pre>*Angel样式:<pre><code>future signature (costume), yellow clothes, yellow hair bow, yellow footwear</code></pre>您可以替换上述标签中的色彩，经测试有效。示例图的LoRA权重为0.8-0.85，您也可以尝试其他权重值。我使用的是WebUI的可选附加网络（additional network）插件，所以示例图的prompt里并没有&lt;lora:XXX:X&gt;的tag，还请注意。画面中有多个主体时可能会出现颜色污染和装饰品缺失的问题如果您愿意使用本LoRA模型并将作品分享给其他人，本人将不胜感激。本人制作的所有模型和绘画皆免费提供，如果觉得还不错的话，欢迎在<a target="_blank" rel="ugc" href="https://www.patreon.com/Kaorin_plkm160">Patreon</a>或<a target="_blank" rel="ugc" href="https://afdian.net/a/KaorinP_plkm160">爱发电</a>上支持我，非常感谢。

フューチャー・シグネチャー | MILLION LIVE!/ミリオンライブ/ミリシタ

V1.0_pynoise: Initial release. Model trained using DAdaptAdam optimizer, with multi-resolution noise and minimal SNR gamma applied.

Files with SHA256: ec7be230b282be69504bf0f806619fab82f0dbbd87744754723198c9f1436a2d