<h2 id="heading-894">中文</h2><p>经历过三次失败才获得这个模型。虽然不是那么完美，但也不错~"yutu"是触发词，添加"hair ornament,long hair,white leotard,white legwear,armored boots,armored gloves,puffy sleeves"获得更好的输出。推荐权重0.8，换服装适当降低权重到0.55左右同时保留触发词和头发相关提示词以获得她的头。可惜的是她的头饰并不能很好的表达出来。目前存在的问题是默认服装在"standing,from behind"的情况下她的屁帘会生成在前面，可以降低权重到0.7放弃这个难用的帘子 :( 感谢使用。训练素材为纯黑背景，使用时添加背景描述词。</p><h2 id="heading-895">English</h2><p>It took three failures to get this model. It's not perfect, but it's not bad~ "yutu" is the trigger word, add "hair ornament, long hair, white leotard, white legwear, armored boots, armored gloves, puffy sleeves" to get better output. The recommended weight is 0.8, and the weight is appropriately reduced to about 0.55 when changing clothes, while retaining the trigger words and hair-related prompt words to get her head. It's a pity that her headgear doesn't express it well. The current problem is that when the default clothing is "standing, from behind", her ass curtain will be generated in front, you can reduce the weight to 0.7 to give up this difficult curtain :( Thanks for using.The training material is a pure black background, and background descriptors are added when using it.</p><h1 id="heading-896">不允许商业用途</h1><h1 id="heading-897">NO COMMERCIAL USAGE</h1>

Files with SHA256: 60463b46f3c286002c0dfc100da48d9d1478ab63180e34e161d1eabfa148aea4