<h2>中文介绍</h2><h3 id="male-capture-panel-krea2-lora">Male Capture Panel｜男性捕捉档案面板 Krea2 LoRA</h3>这是一款用于生成“怪物捕捉 RPG 人物档案界面”的概念 LoRA。画面通常包含一名写实成年男性，以及完整的游戏式人物面板。训练集包含 180 张独立图片，人物主要为 18–22 岁的健康青年男性，覆盖东亚男性与白人男性、校园角色、运动员、技术人员及其他日常身份。面板可以表现：<ul><li>姓名、类型和捕捉状态</li><li>年龄、身高、体重和健康状态</li><li>鞋码、当前心情、职位和出现地点</li><li>捕捉率或稀有度</li><li>体力、速度、力量、魅力、卫生、脚臭等可视化属性条</li><li>位于底部的 <code>SPECIAL NOTE</code> 人物特质介绍</li></ul>训练 Caption 对画面中的英文、文字所在区域、属性条和人物备注进行了详细描述，因此模型不仅能学习面板构图，也能学习“带文字的人物游戏档案”这一整体形式。触发词：<code>malecapturepanel</code>推荐权重：<code>0.65–0.9</code>，建议从 <code>0.8</code> 开始。英文提示词示例：<pre><code>malecapturepanel. Caption: A photorealistic young Chinese male soccer player is displayed inside a complete monster-catching RPG character dossier. The right information panel shows his name, type, age, height, weight, health, shoe size, current mood, position and found-at location. The lower section contains segmented attribute meters for stamina, sprint speed, confidence and foot odor. A detailed SPECIAL NOTE appears inside the bottom text box.</code></pre>控制文字的写法：<pre><code>The upper-right panel reads "NAME: LI HUA", "TYPE: SOCCER", "STATUS: CAPTURE READY", "AGE: 21", "HEIGHT: 175 CM", "WEIGHT: 70 KG", "HEALTH: HEALTHY", "SHOE SIZE: EU 40" and "CURRENT MOOD: HAPPY".</code></pre>推荐负面提示：<pre><code>collage, contact sheet, multiple panels, duplicated person, cropped interface, unreadable text, distorted letters, broken UI, malformed attribute bars</code></pre>注意事项：该 LoRA 能明显提高面板结构、字段布局和英文标签的生成稳定性，但生成模型仍可能拼错姓名或长句。若希望文字更准确，建议使用短姓名、常见英文单词、简短数值，并避免一次要求过长的备注。<hr /><h2 id="english-description">English Description</h2><h3 id="male-capture-panel-or-krea2-concept-lora">Male Capture Panel | Krea2 Concept LoRA</h3>Male Capture Panel is a concept LoRA for creating photorealistic adult men inside detailed monster-catching RPG character dossiers.The dataset contains 180 independent images featuring healthy young adult men, primarily between 18 and 22 years old. It includes East Asian and white male subjects portrayed as college students, athletes, technicians, campus workers and other everyday character classes.The generated interface can include:<ul><li>Name, type and capture status</li><li>Age, height, weight and health</li><li>Shoe size, current mood, position and found-at location</li><li>Catch rate or rarity</li><li>Visual attribute meters for stamina, speed, strength, charisma, hygiene, foot odor and other traits</li><li>A detailed <code>SPECIAL NOTE</code> box at the bottom</li></ul>Every training image uses a detailed natural-language Caption describing the visible English text, its location inside the interface, the attribute meters and the character note. This helps the LoRA learn both the panel composition and the visual language of a text-heavy game dossier.Trigger:<code>malecapturepanel</code>Recommended weight:<code>0.65–0.9</code>; start at approximately <code>0.8</code>.Prompt example:<pre><code>malecapturepanel. Caption: A photorealistic young white male hockey player appears inside a complete monster-catching RPG capture dossier. His portrait occupies the left side, while the right panel displays his name, type, status, age, height, weight, health, shoe size, current mood, position and location. Segmented meters below show stamina, capture resistance, hygiene and foot odor. A multi-sentence SPECIAL NOTE appears in the bottom information box.</code></pre>Example for controlling the visible text:<pre><code>The upper-right information panel reads "NAME: BLAKE MERCER", "TYPE: ICE CHARGER", "STATUS: WILD", "AGE: 21", "HEIGHT: 186 CM", "WEIGHT: 84 KG", "HEALTH: EXCELLENT" and "SHOE SIZE: US 11".</code></pre>Suggested negative prompt:<pre><code>collage, contact sheet, multiple panels, duplicated person, cropped interface, unreadable text, distorted letters, broken UI, malformed attribute bars</code></pre>Limitations:The LoRA improves interface structure, field placement and English-label consistency, but exact spelling is not guaranteed. Short names, familiar English words, simple values and concise special notes generally produce the best results.

Files with SHA256: 9d07eb6900fb0d9be4c874d064c28301788eeb026cc7660db5417d36d9317e0e