This LoRA model is a character enhancement model for the adult version of Cyrene, fine-tuned on a dataset of 500+ images.Special optimization has been applied to address the hard-baked clothing issue. Cyrene's official accessories are now detachable: the headdress is primarily bound to "hair ornament", the dress can be modified/removed using "white dress" and "sleeveless dress", the head veil is bound to "veil", and her shoes are bound to "white footwear". Omitting these clothing tags will default to her official illustration appearance. During testing with my usual Illustrious series checkpoint model, removing these tags allows for adding any preferred clothing elements. If residual elements persist when using your checkpoint for inference, consider adding the aforementioned phrases to the negative prompt.<ul><li>Recommended weight: 0.7-1.0</li><li>Example prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, cute, playful, romantic, love, dynamic angle,
</code></pre><ul><li>Full body prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, white footwear, full body,</code></pre>While I am very satisfied with the accuracy of Cyrene's features,including her distinctive diamond-shaped pupils, headwear shape, and breastplate (features not consistently achieved in other LoRAs I've tested), I plan to double the epochs for retraining after the official 3D model is released.In addition to the version based on Illustrious XL, I have also trained a version based on ZukiAnimeIll (a derivative model of Illustrious XL). Users employing ZukiAnimeIll for inference can use either this version or that specific one.(Please adjust the prompts according to the official documentation of the checkpoint model you are using.)If you have any feedback, please feel free to leave comments so I can make adjustments. Thank you.这个LoRA模型是成人形态昔涟的角色增强模型，在500+图片的数据集上微调得到。这次特意对角色服装hard-baked问题进行了优化，现在昔涟的官方外设是可拆卸的，头饰基本绑定在”hair ornament”，裙子可以删改”white dress”和”sleeveless dress”，头帘绑定在”veil”，昔涟的鞋子绑定在”white footwear”里。不加服装提示词就会默认呈现官方立绘造型。我在使用我常用的Illustrious系列checkpoint模型进行推理测试，在去掉这些提示词后可以添加任意喜欢的服装元素，如果你在使用你的checkpoint模型推理的时候仍有残留，可以考虑在负面提示词一栏加入上述短语。<ul><li>推荐权重：0.7-1.0</li><li>示例prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, cute, playful, romantic, love, dynamic angle,
</code></pre><ul><li>全身prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, white footwear, full body,
</code></pre>（提示词请根据你使用的checkpoint模型的官方文档进行调整）虽然我已经对昔涟的特征还原程度很满意，包括她独特的菱形瞳孔，头饰形状，胸甲等（目前我试用过其他的LoRA没有在这些特征做到一致），但我会在官方3d模型发布后翻倍epoch再训练。除了基于Illustrious XL的版本，我还训练了基于ZukiAnimeIll（Illustrious XL的一款衍生模型）的版本，使用ZukiAnimeIll推理的用户可以使用该版，也可以使用那一版。如果有任何反馈意见欢迎评论，以便我做出调整，谢谢。

Cyrene|大昔涟 （ultra detailed 细节增强）Honkai:Star Rail 崩坏：星穹铁道

This LoRA model is a character enhancement model for the adult version of Cyrene, fine-tuned on a dataset of 500+ images. (This version is specified for ZukiAnimeIll)Special optimization has been applied to address the hard-baked clothing issue. Cyrene's official accessories are now detachable: the headdress is primarily bound to "hair ornament", the dress can be modified/removed using "white dress" and "sleeveless dress", the head veil is bound to "veil", and her shoes are bound to "white footwear". Omitting these clothing tags will default to her official illustration appearance. During testing with my usual Illustrious series checkpoint model, removing these tags allows for adding any preferred clothing elements. If residual elements persist when using your checkpoint for inference, consider adding the aforementioned phrases to the negative prompt.<ul><li>Recommended weight: 0.7-1.0</li><li>Example prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, cute, playful, romantic, love, dynamic angle,
</code></pre><ul><li>Full body prompt:<pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, white footwear, full body,
</code></pre></li></ul>While I am very satisfied with the accuracy of Cyrene's features,including her distinctive diamond-shaped pupils, headwear shape, and long braided hair (features not consistently achieved in other LoRAs I've tested), I plan to double the epochs for retraining after the official 3D model is released.(Please adjust the prompts according to the official documentation of the checkpoint model you are using.)If you have any feedback, please feel free to leave comments so I can make adjustments. Thank you.这个LoRA模型是成人形态昔涟的角色增强模型，在500+图片的数据集上微调得到。（这个版本的LoRA是基于ZukiAnimeIll的）这次特意对角色服装hard-baked问题进行了优化，现在昔涟的官方外设是可拆卸的，头饰基本绑定在”hair ornament”，裙子可以删改”white dress”和”sleeveless dress”，头帘绑定在”veil”，昔涟的鞋子绑定在”shoes”里。不加服装提示词就会默认呈现官方立绘造型。我在使用我常用的Illustrious系列checkpoint模型进行推理测试，在去掉这些提示词后可以添加任意喜欢的服装元素，如果你在使用你的checkpoint模型推理的时候仍有残留，可以考虑在负面提示词一栏加入上述短语。<ul><li>推荐权重：0.7-1.0</li><li>示例prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, cute, playful, romantic, love, dynamic angle,
</code></pre><ul><li>全身prompt:</li></ul><pre><code>(masterpiece, best quality, very aesthetic), Cyrene, 1girl, long hair, gradient hair, solo, white dress, looking at viewer, hair ornament, white dress, sleeveless dress, white footwear, full body,
</code></pre>（提示词请根据你使用的checkpoint模型的官方文档进行调整）虽然我已经对昔涟的特征还原程度很满意，包括她独特的菱形瞳孔，头饰形状，麻花辫等（目前我试用过其他的LoRA没有在这些特征做到一致），但我会在官方3d模型发布后翻倍epoch再训练。如果有任何反馈意见欢迎评论，以便我做出调整，谢谢。

Files with SHA256: 02f44deead9802dd1ab1893c279ed40f40212984445ab28df58b8e6a9831ec29