Azure VerlunaMix

Verluna Core Stable Update Changelog

Verluna Core Stable is an updated refinement of the Verluna.This version focuses on delivering a cleaner, more controlled, and more stylized

🔧 What’s New in This Update

Improved Generation Stability
Output is now more consistent with fewer unpredictable artifacts.
More Accurate Prompt Response
The model follows keywords, style cues, and character prompts more reliably.
Dataset Upgrade: E6 Dataset
Updated dataset improves species diversity, texture consistency, and anatomical understanding.
Reduced Background Clutter
Backgrounds remain present but are cleaner and less chaotic, improving composition clarity.
Softer Character Rendering (Toony Style)
Characters now generate with a softer and slightly more stylized look, leaning toward a light toon aesthetic.

Azure VerlunaMix 是一个以毛茸茸角色为核心的基座模型，专为平衡的解剖结构、动态构图和增强的半写实纹理而设计。
该模型融合了多个专业检查点，可在各种毛茸茸物种、光照和环境条件下提供视觉丰富且高度适应的渲染性能。

模型融合组成

🐾 Mol_Keun Mix - DeepBlue V2 EPS | NoobAI 检查点 | Civitai
🦊 Banana Splitz XL - Vee 5 Point Oh! | Illustrious 检查点 | Civitai
❄️ Azure ColdSnap - ColdSnap | Illustrious 检查点 | Civitai
✨ + 2 个私有 LoRa 风格融合

⚙️ 推荐设置

步数： 25–40
CFG 缩放： 4–8

⚠️ 若超过 9，根据提示的密集程度，可能导致过曝或伪影。

采样器： Euler a / DPM++ 2M Karras（用户偏好）
高分辨率修复： ✅ 推荐
FaceDetailer（ComfyUI）： 适用于特写或肖像场景，可选

🏷️ 提示标签指南

✅ 正向标签（提升质量）

masterpiece, amazing quality, best quality, ultra-detailed, 8K, illustration, CG, ultra-detailed fur, detailed eyes, detailed background, highres, very awa, cinematic lighting, dynamic pose, newest, clean composition

🚫 负向标签（建议使用）

lowres, (worst quality:1.2), (bad quality:1.2), (low quality:1.2), bad anatomy, bad hands, fused fingers, too many fingers, old, extra limb, mutated body, poorly drawn fur

模型类型	检查点
基础模型	Illustrious
发布时间	11/23/2025

Azure VerlunaMix

详情

下载文件

关于此版本

模型描述