请不要使用该模型用于商业用途。Please do not use this model for commercial purposes.请注意选择版本使用。Please pick the appropriate version to use this LoRa.以下内容使用机器翻译 The following uses machine translation（To English)以下所描述内容非专业分析，仅为个人使用经验。The content described below is not a professional analysis and is only a summary of personal experience.由于素材较少，难免出现重复的动作或者有明显的风格倾向，请降低采样步数和Lora权重使用。Due to the limited amount of material, repetitive actions or obvious style tendencies are inevitable. Please reduce the number of sampling steps and Lora weights for use.范例图虽然完全提供了生成该图片时的全部Tag，但请注意有些版本的LoRa名称在我上传时进行了改动Although the example image provides all the tags at the time the image was generated, please note that some versions of LoRa have changed their names when I uploaded them.角色原型来自游戏：Blue ArchiveThe character prototype comes from the game: Blue Archive如果使用该LoRa去生成裸露的或者其他非安全的图片，可能获得效果不尽人意。If you use LoRa to generate nude or other nsfw, you may not be as effective as you should be.一些范例图我使用的是自用模型，其基于众多模型的融合，但主体是AnythingV4.5。Some example diagrams I use my own model, which is based on the fusion of many models, but the main body is AnythingV4.5. 用合适的tag去描述符合角色原本的年龄、性别特征可能获得更好的成果。Using appropriate tags to describe the original age and gender characteristics of the character may obtain better results.这种LoRa应该没有提示词，请直接添加&lt;lora:HanaokaYuzuV009:1&gt; or&lt;lora:HanaokaYuzuV010:1&gt;or&lt;lora:HanaokaYuzuV011:1&gt;or&lt;lora:HanaokaYuzuV012:1&gt;or&lt;lora:HanaokaYuzuV012p:1&gt;（根据版本）使用。This LoRa should have no prompt, please add &lt;lora:HanaokaYuzuV009:1&gt; or&lt;lora:HanaokaYuzuV010:1&gt;or&lt;lora:HanaokaYuzuV011:1&gt;or&lt;lora:HanaokaYuzuV012:1&gt;or&lt;lora:HanaokaYuzuV012p:1&gt;（as version）to tags for use it.如果认为范例图的紫色眼睛不是很还原，可以尝试直接描述眼睛的颜色,例如： grey eyes，或者不描述。If you think purple eyes are not very reductive, try to describe the color of the eyes directly：graw eyes. Or not described.V009<ul><li>这个版本是我早期用来生成更多训练素材时使用的，如果你认为其他版本不好控制，可以尝试使用该版本。This version was used earlier when I used to generate more training material, and if you think other versions are not easy to control, you can try this version。</li></ul>V010推荐使用的Tag（不包含通用的起手tag）/Recommended tag(Does not include the generic base start tag)：<ul><li>通用/General&lt;lora:HanaokaYuzuV010:0.7&gt;,red_hair,small_breasts,long_hair,very_long_hair</li><li>如果认为画面饱和度过高，请进一步降低权重使用If you think that the picture is too saturated, please further reduce the weight usage</li><li>尝试出现光环 Try to make a halo appear in the picturehalo</li><li>尝试出现长发上的蝴蝶结发饰/Try to makehair bow on hairhair bow or flower</li><li>常服 /clotheshood,jacket,white_jacket</li><li>如果连续出现同一姿势的情况，请尝试在负面提示词中使用以下tag/If the same pose occurs in succession, please try to add the following tag in the Negative prompt:tachi-e, tachi-e_stand,</li><li>如果连续出现同一姿势的情况，请避免使用以下tag/If the same pose occurs in succession, avoid using the following tagtachi-e, tachi-e_stand,</li></ul>V011<ul><li>优化了饱和度和明度过高的问题。</li><li>Optimized the problem of high saturation and brightness.</li></ul>V012<ul><li>意外的优化了饱和度和明度过高的问题。Unexpected optimization of saturation and over-brightness problems.</li><li>增加了女仆装/Add Enmaided女仆装推荐使用的Tag/Recommended tag to take Enmaided：apron, gradient, maid, white_apron, black_dress, ponytail, black_footwear, frilled_apron, dress, maid_apron, alternate_hair,very long hair, split_ponytail,</li><li>如果连续出现同一姿势的情况，请尝试在负面提示词中使用以下tag/If the same pose occurs in succession, please try to add the following tag in the Negative prompt:tachi-e, tachi-e_stand,</li><li>如果连续出现同一姿势的情况，请避免使用以下tag/If the same pose occurs in succession, avoid using the following tagtachi-e, tachi-e_stand, holding_a_weapon, alternate_weapon,gradient_background, gradient, blue_background,</li></ul>V012p<ul><li>加多了V012对女仆装的学习步数/Added the number of steps for V012 to the maid outfit</li><li>增加了对女仆装的一些特征的标注/Added annotations for some features of maid</li><li>tag请参照范例图/Please refer to the example diagram for tags</li></ul>

Hanaoka yuzu 花岡ユズ 花冈柚子 Blue Archive

HanaokaYuzuV012p.safetensors

Files with SHA256: adb116a7773396206dda23f278b67b0ef55adb8ac7a5a30d99249cca3802d7bf