<h2>Created with Tutu AI Workflow</h2>
<p><strong>Tagged with Tutu Annotator. Trained with Tutu Trainer.</strong></p>
<p>This LoRA is part of my Tutu AI workflow. Tutu Trainer is used for LoRA training, model resource management, parameter presets, and training checks. Tutu Annotator is used for dataset annotation, prompt reverse engineering, English-friendly tags, and cleaner searchable captions.</p>
<p>For this Z-Image-Turbo experiment, the source material was annotated with the Tutu intelligent labeling workflow, including Banana 2 / Banana PRO-style visual annotation and subject-aware filtering. The goal is to keep the captions focused on the outfit shape, material feeling, and accessory structure instead of unrelated background details.</p>
<p><img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/b87f65a3-f843-47b2-b754-cc37264ff1c2/width=1600/b87f65a3-f843-47b2-b754-cc37264ff1c2.png" alt="Tutu AI Workflow - Tutu Trainer and Tutu Annotator" /></p>
<ul>
<li><strong>Official website:</strong> <a target="_blank" rel="ugc" href="https://zhaotutu.xyz/">https://zhaotutu.xyz/</a></li>
<li><strong>Tutu Trainer:</strong> <a target="_blank" rel="ugc" href="https://zhaotutu.xyz/downloads/tututrainer/">https://zhaotutu.xyz/downloads/tututrainer/</a></li>
<li><strong>Tutu Annotator:</strong> <a target="_blank" rel="ugc" href="https://zhaotutu.xyz/downloads/tutuannotator/">https://zhaotutu.xyz/downloads/tutuannotator/</a></li>
<li><strong>Bilibili:</strong> <a target="_blank" rel="ugc" href="https://space.bilibili.com/431046154">https://space.bilibili.com/431046154</a></li>
<li><strong>YouTube:</strong> <a target="_blank" rel="ugc" href="https://www.youtube.com/@zhaotutu/videos">https://www.youtube.com/@zhaotutu/videos</a></li>
<li><strong>Telegram:</strong> <a target="_blank" rel="ugc" href="https://t.me/zhaotutu">https://t.me/zhaotutu</a></li>
<li><strong>TikTok:</strong> <a target="_blank" rel="ugc" href="https://www.tiktok.com/@zhaotututu">https://www.tiktok.com/@zhaotututu</a></li>
<li><strong>Reddit:</strong> <a target="_blank" rel="ugc" href="https://www.reddit.com/user/zhaotututu/">https://www.reddit.com/user/zhaotututu/</a></li>
<li><strong>GitHub:</strong> <a target="_blank" rel="ugc" href="https://github.com/zhaotututu">https://github.com/zhaotututu</a></li>
<li><strong>Email:</strong> bigpox12@gmail.com</li>
<li><strong>QQ:</strong> 331506796</li>
</ul>

<h2>Tutu HiSilk Z-Image-Turbo: Black Accessories Jumpsuit</h2>
<p>This is an experimental <strong>Z-Image-Turbo HiSilk LoRA</strong> for a black tight-fitting jumpsuit with accessory details. It focuses on a smooth nylon-like black outfit, close-fitting bodysuit / jumpsuit structure, black high heels, wide strap details, metal chain accents, and a futuristic fashion styling direction.</p>
<p>This is not the final form of the Z-Image direction. The model is already usable, but it still has room for improvement. The base model already understands many clothing concepts, so this LoRA mainly helps steer the outfit style, material impression, and accessory structure.</p>
<ul>
<li><strong>Base model:</strong> Z-Image-Turbo / ZImageTurbo</li>
<li><strong>Type:</strong> black accessories jumpsuit LoRA</li>
<li><strong>Trigger word:</strong> <code>tutuliantiyi</code></li>
<li><strong>Recommended LoRA weight:</strong> around <strong>1.0</strong></li>
<li><strong>Loader:</strong> normal LoRA loader</li>
<li><strong>Compatibility:</strong> compatible with Z-Image-Turbo workflows</li>
</ul>

<h3>What this model is good at</h3>
<ul>
<li>Adding a black tight-fitting jumpsuit / bodysuit direction to Z-Image-Turbo generations.</li>
<li>Strengthening smooth nylon-like material and HiSilk-style clothing texture.</li>
<li>Introducing strap, chain, buckle, and accessory details without changing the whole workflow.</li>
<li>Working with normal LoRA loading methods.</li>
<li>Serving as an early Z-Image-Turbo test for the HiSilk outfit series.</li>
</ul>

<h3>Training notes</h3>
<p>This model was made from photo source material. The dataset was labeled with Tutu Annotator using Banana 2 / Banana PRO-style visual annotation, then filtered with Tutu's subject-aware tag filtering. The training was completed with Tutu Trainer and the Tutu training workflow.</p>
<p>The purpose of this experiment is to test how a specific outfit and accessory concept transfers into Z-Image-Turbo. Different Z-Image workflows may need slightly different LoRA weights and prompt strength.</p>

<h3>Recommended use</h3>
<ul>
<li>Use a Z-Image-Turbo compatible workflow.</li>
<li>Load this LoRA with a normal LoRA loader.</li>
<li>Start with LoRA weight around <strong>1.0</strong>.</li>
<li>Put the trigger word <code>tutuliantiyi</code> near the beginning of the prompt.</li>
<li>If the outfit is weak, strengthen the black jumpsuit, bodysuit, straps, chains, and accessory descriptions.</li>
<li>If the outfit becomes too dominant, lower the LoRA weight first.</li>
</ul>

<h3>Prompt keywords</h3>
<p><strong>Main trigger:</strong></p>
<p><code>tutuliantiyi</code></p>
<p><strong>Useful prompt phrases:</strong></p>
<ul>
<li><code>black tight-fitting jumpsuit</code></li>
<li><code>black bodysuit</code></li>
<li><code>smooth nylon-like material</code></li>
<li><code>wide black straps</code></li>
<li><code>metal chain accessories</code></li>
<li><code>black high heels</code></li>
<li><code>futuristic fashion outfit</code></li>
<li><code>realistic clothing texture</code></li>
</ul>
<p>Example prompt fragment: <code>tutuliantiyi, wearing a black tight-fitting jumpsuit and black high heels, smooth nylon-like material, wide black straps, metal chain accessories, futuristic fashion outfit</code>.</p>

<h3>Z-Image workflow</h3>
<p>This LoRA is compatible with many Z-Image-Turbo workflows. My workflow resources are still being expanded here:</p>
<p><a target="_blank" rel="ugc" href="https://www.runninghub.ai/user-center/1936823199386537986/webapp?inviteCode=rh-v0990">RunningHub workflow page</a>. New users can use invite code <strong>rh-v0990</strong> if needed.</p>

<h3>Responsible use</h3>
<ul>
<li>You are fully responsible for any creative work made with this model.</li>
<li>Do not use this model to create or share illegal, harmful, malicious, defamatory, fraudulent, or political content.</li>
<li>If you use this model commercially or need custom LoRA cooperation, please contact me first. Thank you.</li>
</ul>

<h2>中文说明</h2>
<p>这是 <strong>图图嗨丝造相 - Z-Image-Turbo 黑色配饰连体衣 LoRA</strong> 的实验版本。它主要用于黑色紧身连体衣 / bodysuit 方向，包含光滑尼龙感材质、黑色高跟鞋、宽带、链条、扣环等配饰细节，以及偏未来感的时装造型。</p>
<p>这还不是 Z-Image 方向的最终形态。模型已经可以使用，但仍然有提升空间。由于底模本身已经理解很多服装概念，这个 LoRA 主要用于稳定服装款式、材质感觉和配饰结构。</p>
<ul>
<li><strong>基础模型：</strong>Z-Image-Turbo / ZImageTurbo</li>
<li><strong>类型：</strong>黑色配饰连体衣 LoRA</li>
<li><strong>触发词：</strong><code>tutuliantiyi</code></li>
<li><strong>推荐 LoRA 权重：</strong><strong>1.0</strong> 左右</li>
<li><strong>加载方式：</strong>正常 LoRA 加载器即可</li>
<li><strong>兼容性：</strong>兼容各种 Z-Image-Turbo 工作流</li>
</ul>

<h3>这个模型适合什么</h3>
<ul>
<li>在 Z-Image-Turbo 工作流里加入黑色紧身连体衣 / bodysuit 方向。</li>
<li>强化光滑尼龙感、嗨丝材质感和贴身服装质感。</li>
<li>增加宽带、链条、扣环等配饰结构。</li>
<li>使用普通 LoRA 加载方式即可加载。</li>
<li>作为嗨丝服装系列在 Z-Image-Turbo 上的早期实验版本。</li>
</ul>

<h3>训练说明</h3>
<p>这个模型使用照片素材制作。素材用 Tutu Annotator / 图图打标器进行标注，使用 Banana 2 / Banana PRO 视觉标注能力辅助打标，并通过图图智能主体过滤功能过滤无关标签。训练使用 Tutu Trainer / 图图训练器和图图训练工作流完成。</p>
<p>这个实验的重点，是测试特定服装和配饰概念在 Z-Image-Turbo 里的迁移效果。不同 Z-Image 工作流可能需要略微调整权重和提示词强度。</p>

<h3>推荐用法</h3>
<ul>
<li>使用兼容 Z-Image-Turbo 的工作流。</li>
<li>用正常 LoRA 加载器加载。</li>
<li>LoRA 权重从 <strong>1.0</strong> 左右开始。</li>
<li>在提示词开头附近加入触发词 <code>tutuliantiyi</code>。</li>
<li>如果服装效果不明显，请加强黑色连体衣、紧身衣、宽带、链条和配饰描述。</li>
<li>如果服装效果过强，先降低 LoRA 权重。</li>
</ul>

<h3>出图关键词</h3>
<p><strong>主触发词：</strong></p>
<p><code>tutuliantiyi</code></p>
<p><strong>可用提示词：</strong></p>
<ul>
<li><code>black tight-fitting jumpsuit</code></li>
<li><code>black bodysuit</code></li>
<li><code>smooth nylon-like material</code></li>
<li><code>wide black straps</code></li>
<li><code>metal chain accessories</code></li>
<li><code>black high heels</code></li>
<li><code>futuristic fashion outfit</code></li>
<li><code>realistic clothing texture</code></li>
</ul>
<p>提示词片段示例：<code>tutuliantiyi, wearing a black tight-fitting jumpsuit and black high heels, smooth nylon-like material, wide black straps, metal chain accessories, futuristic fashion outfit</code>。</p>

<h3>Z-Image 工作流</h3>
<p>这个 LoRA 兼容多种 Z-Image-Turbo 工作流。我的工作流资源还在这里逐步补充：</p>
<p><a target="_blank" rel="ugc" href="https://www.runninghub.ai/user-center/1936823199386537986/webapp?inviteCode=rh-v0990">RunningHub 工作流页面</a>。新人如果需要，可以使用邀请码 <strong>rh-v0990</strong>。</p>

TuTu's HiSilk 造相 - Z-Image-Turbo (Black Accessories Jumpsuit)/图图的嗨丝造相-Z-Image-Turbo（黑色配饰连体衣）

<p><strong>v1.0 - Z-Image-Turbo experimental version.</strong></p>
<p>Trigger word: <code>tutuliantiyi</code>. Recommended LoRA weight: around <strong>1.0</strong>. Load it with a normal LoRA loader in a Z-Image-Turbo compatible workflow.</p>
<p>This version focuses on a black tight-fitting jumpsuit / bodysuit direction with smooth nylon-like material, black high heels, wide strap details, chain accessories, and futuristic fashion styling.</p>
<p>If the outfit effect is weak, strengthen the prompt with phrases such as <code>black tight-fitting jumpsuit</code>, <code>black bodysuit</code>, <code>wide black straps</code>, and <code>metal chain accessories</code>. If the outfit becomes too dominant, lower the LoRA weight first.</p>
<p><strong>中文：</strong>这是 Z-Image-Turbo 实验版本。触发词 <code>tutuliantiyi</code>，推荐权重 <strong>1.0</strong> 左右，使用正常 LoRA 加载器即可。主要方向是黑色紧身连体衣 / bodysuit、光滑尼龙感材质、宽带、链条和配饰结构。如果效果不明显，请加强服装与配饰描述；如果过强，优先降低权重。</p>

Files with SHA256: 286c03526da085d441de50f3dac1b04d2024a593c18b70aa6df628b6b05179e0