<h2 id="bunny-3d-gamevision">🐰 BUNNY 3D GameVision</h2><h2 id="krea2-update">KREA2 Update</h2><ul><li>No trigger word required. 无需触发词。</li><li>Recommended weight: 1.0–1.2. 推荐权重：1.0–1.2。</li><li>Works best when combined with 3D-related prompts, which can push the effect further. 配合3D相关提示词时效果更强，更容易拉出建模感。</li><li>For already-3D character LoRAs, this works more like an enhancer, improving model feel, materials, and overall presentation without heavily changing the character. 对于本身就是3D角色的人物 LoRA，它更像是一个增强器，主要提升建模感、材质和整体完成度，而不会大幅改掉角色本身。</li><li>On some base models that are specifically tuned for realism enhancement, the effect may be less obvious and behave more like a subtle refinement. 在某些专门强化现实感/写实感的底模上，效果可能不会特别强，更像是一种轻度微调。<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/211dea59-b206-45e9-b787-c8074752e111/original=true/211dea59-b206-45e9-b787-c8074752e111.jpeg" /></li></ul>A no-trigger AAA game CG style LoRA for ANIMA, designed to strengthen 3D rendering, materials, lighting, and character detail while preserving facial identity.一款无需触发词的 ANIMA 3A游戏CG风格 LoRA，用于增强3D渲染、材质、光影和人物细节，同时尽量保持人物面部身份<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/867172f7-ab39-4b2d-9bc6-02b92a125073/original=true/867172f7-ab39-4b2d-9bc6-02b92a125073.jpeg" /><h2 id="about">About</h2>This LoRA was trained on 360 manually selected images from 48 different 3D games, covering characters, armor, clothing, environments, props, vehicles, and action scenes. 本 LoRA 使用了来自 48 款不同3D游戏的360张人工筛选图片进行训练，内容涵盖人物、盔甲、服装、环境、道具、载具和动作场景。Its purpose is to push ANIMA and compatible character LoRAs toward a more consistent 3D game CG appearance without requiring additional style prompts. 它的目标是让 ANIMA 以及兼容的人物 LoRA 在不需要额外风格提示词的情况下，更稳定地呈现出 3D游戏CG效果。It mainly improves material rendering, lighting, facial volume, clothing structure, armor surfaces, environmental depth, and overall model-like presentation. 它主要增强材质表现、光影、面部体积、服装结构、盔甲表面、环境纵深以及整体建模感。Most importantly, during my tests with multiple character LoRAs, the character’s facial identity remained stable while the overall 3D presentation became stronger. 最重要的是，在我使用多个人物 LoRA 测试时，人物面部身份能够保持稳定，同时整体3D表现得到了明显增强。<hr /><h2 id="usage">Usage</h2>No trigger word is required. 无需触发词。Simply load it together with your ANIMA character LoRA. 直接与 ANIMA 人物 LoRA 一起加载即可。<h3 id="recommended-weight">Recommended Weight</h3><ul><li>0.8: Recommended default weight for most character LoRAs. 0.8： 大多数人物 LoRA 的默认推荐权重。</li><li>0.9–1.0: Stronger 3D conversion, richer materials and modeling. 0.9–1.0： 3D化、材质和建模感更强。</li></ul><hr /><h2 id="compatibility">Compatibility</h2>This LoRA works best with 3D, semi-realistic, game-character, or materially complex character LoRAs, especially my <a target="_blank" rel="ugc" href="https://civarchive.com/collections/17064201">BUNNY character series.</a> 本 LoRA 更适合 3D、半写实、游戏人物以及服装材质较复杂的人物 LoRA，尤其适合我的 <a target="_blank" rel="ugc" href="https://civarchive.com/collections/17064201">BUNNY 人物系列</a>。It can help reveal clothing structures, accessories, tattoos, armor details, and other information that the original ANIMA output may not fully express. 它可以帮助表现原始 ANIMA 出图中没有被充分释放的服装结构、配饰、纹身、盔甲细节等人物信息。For heavily stylized, flat, or strongly cartoon-like character LoRAs, the 3D conversion may be less noticeable because the character LoRA’s original visual language remains dominant. 对于高度平面化、卡通化或自身风格非常强烈的人物 LoRA，3D化效果可能不会特别明显，因为人物 LoRA 本身的视觉特征仍然占据主导。<hr /><h2 id="notes">Notes</h2>This is not intended to completely redesign a character or replace the character LoRA’s identity. 它的目的不是重新设计角色，也不是取代人物 LoRA 本身的身份特征。The character LoRA provides the face, clothing, accessories, and character-specific details, while this LoRA provides the underlying 3D game CG presentation. 人物 LoRA 负责面部、服装、配饰和角色专属细节，而本 LoRA 负责提供底层的3D游戏CG表现。

Files with SHA256: 10d062eba2d7c1f5e78622ac6da3c7ee48f1f0a77d25736b98348833100fbfad