XL_v2.1模型使用精简优化的100+张图片作为训练集，并增加了几个角色的标签绑定，整体效果更接近藤岛康介手绘复古风格。训练图片主要来自藤岛康介的原画集，还有一部分来自动画版原画师：松原秀典。FLUX版本：<a target="_blank" rel="ugc" href="https://civarchive.com/models/722342">https://civarchive.com/models/722342</a>illustriousXL 版本：<a target="_blank" rel="ugc" href="https://civarchive.com/models/98790">https://civarchive.com/models/98790</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/907234">4</a>Animagine XL V3 和 SD1.5 版本：<a target="_blank" rel="ugc" href="https://civarchive.com/models/73822">https://civarchive.com/models/73822</a>--- --- --- --- --- --- --- ---The XL_v2.1 model uses 100 streamlined and optimized images as the training set, and adds Trigger Words for several characters.The overall effect is closer to the hand-painted retro style of Kosuke Fujishima.The training pictures are mainly from the original painting collection of Kosuke Fujishima, and some are from the original artist of the animation version: Hidenoria Matsubara.FLUX version：<a target="_blank" rel="ugc" href="https://civarchive.com/models/722342">https://civarchive.com/models/722342</a>illustriousXL version：<a target="_blank" rel="ugc" href="https://civarchive.com/models/98790">https://civarchive.com/models/98790</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/907234">4</a>Animagine XL V3 and SD1.5 version：<a target="_blank" rel="ugc" href="https://civarchive.com/models/73822">https://civarchive.com/models/73822</a>

Files with SHA256: b03e57f101b0ebaed43279d42c54213b7feba1210ee07bec4b41beec40bc3dba