<p>[English/中文]</p><p>This is my first LoRA model and there had been 7 previous version trained before the fromal version was released. It will be highly appreciated if you could offer me any advice!</p><h1 id="heading-286">Basic Information</h1><p>Trigger word: arima kana \(oshi no ko\)</p><p>Training model: Anything V5 PrtRE (I suggest you use the same model to generate images)</p><p>Recommended weight: 0.8-1.0</p><h1 id="heading-287">Example Image Introduction</h1><p>The relatively extensive datasets enable more elements to be changeable. You'd better look at example images for proper prompts. The example images (originally 600 * 900 definition, 2x SD upsale) are based on Anything V5 PrtRE and AnimeVAE. I don't use any extension except EasyNegativeV2. The same basic parameters will produce the same result.</p><p>1-2.       high school uniform (Unfortunately, it is hard for AI to learn to the fact that there is a black bowtie on the back of Kana’s head)</p><p>3-5.       a few sets of clothes appeared in the Oshi No Ko anime (not 100% similar)</p><p>6-8.       different expressions (I have to admit that the model has quite limited expressions)</p><p>9-11.      different angles and poses using Controlnet Openpose (When Openpose is used, the result will be less authentic)</p><p>12-13.     other types of clothes</p><p>14-16.   greyscale style and other checkpoint models (Obviously, other checkpoint models are less effective than Anything V5 PrtRE)</p><p>17.          NSFW <s>(What, do you say that Arima Kana only has small breasts?)</s></p><h2 id="heading-288">Author’s Words <s>(you can skip it)</s></h2><p>Arima Kana is a special girl. She has her particular temperament. So I want to produce a LoRA that can paint Arima Kana, not someone who barely has short red hair, bangs and red eyes. It is a quite challengeable task which few present LoRAs (uploaded in CivitAI) can complete.  I spent a lot of time painstakingly selecting qualified datasets to achieve this goal.</p><h1 id="heading-289">Acknowledgement</h1><p>I would be glad if you can like the model. Don’t forget to support me!</p><h3 id="heading-290"><strong>Finally but most importantly, Arima Kana is the best!</strong></h3><p></p><p></p><p>这是我训练的第一个LoRA模型，在最终版本之前已经迭代过7个版本了......欢迎提出宝贵的意见！</p><h1 id="heading-291">基本信息</h1><p>触发词：arima kana \(oshi no ko\)</p><p>训练底模：Anything V5 PrtRE（我建议也用这个模型生成图片）</p><p>推荐权重：0.8-1.0</p><h1 id="heading-292">例图介绍</h1><p>得益于比较多样的训练集，模型产出的图片也将更具多样性。你可以去参考例图的提示词来实现这些效果。例图原始分辨率为600*900，用了2倍的SD放大，底模Anything V5 PrtRE，使用AnimeVAE，并且除了EasyNegativeV2没有使用任何其他额外拓展。相同的基本参数即可产出与例图一样的图片。</p><p>1-2.        高中校服（然而AI其实很难意识到加奈后脑勺还有个蝴蝶结）</p><p>3-5.        一些在推子动漫出现过的服饰（没办法完全复刻）</p><p>6-8.        不同的表情（我承认这个模型的表情并不是很丰富）</p><p>9-11.      用Controlnet Openpose产出的不同角度与姿势（Openpose会导致较低的还原度）</p><p>12-13.     其他服饰</p><p>14-16.     黑白风格和其它大模型（显然，其他模型的产出效果不如Anything V5 PrtRE）</p><p>17.          懂的都懂<s>（什么，你想说帽子没那么大？）</s></p><h2 id="heading-293">作者的话<s>（可跳过）</s></h2><p>有马加奈是一个特别的女孩，有着自己的气质。我希望我做的LoRA可以画出那个特别的有马加奈，而不是随便一个有着红色短发、齐刘海、红瞳的人。这个要求非常难实现，目前的C站上加奈的LoRA都很难做到这点。为此我花了很多时间精心挑选数据集以最大程度还原她的特征。</p><h1 id="heading-294">鸣谢</h1><p>如果你能喜欢这个模型，那就太好啦，同时也别忘了支持我一下噢~</p><h3 id="heading-295"><strong>最后但最重要的是，帽子最高！</strong></h3>

Files with SHA256: 57cb52b4fd8fec7b957468afaa633685284cb58e3ebdee74c173dbeb08d9f4be