# YUNA(์ ๋) is a 100% fictional AI-generated original character (non-existent virtual model).
A finely crafted SDXL character LoRA designed for generating stunning, high-detail female portraits with natural facial features and photorealistic skin textures.
Trained on curated, high-resolution dataset with detailed visual captions, offering exceptional flexibility with clothing, lighting, and background prompts without overfitting.
---
### โ๏ธ Recommended Settings
* Trigger Word: yuna_model
* LoRA Weight: 0.75 - 1 (Sweet spot: 0.9)
* Base Model: Stability AI SDXL Base 1.0
* Compatible Checkpoints:
* RealVisXL V4.0 / V5.0
*HelloWorld XL
* CyberRealistic XL
* Sampling Steps: 30 - 40
* CFG Scale: 3.5 - 6.0
* Sampler: DPM++ 2M Karras / Euler a
* Resolution: 896x1152, 832x1216, or 1024x1024
---
### ๐ก Prompt Guide & Examples
#### ๐ธ Daily Casual / Portrait
> Prompt:
> yuna_model, 1girl, solo, looking at viewer, gentle smile, natural brown hair, white knit sweater, sitting at a sunlit cafe, soft natural lighting, depth of field, photorealistic, 8k uhd, masterpiece
#### ๐ Chic & Elegant / Editorial
> Prompt:
> yuna_model, 1girl, solo, chic expression, elegant black blazer, modern studio background, dramatic rim lighting, sharp focus on eyes, detailed skin texture, fashion magazine photography
#### ๐ซ Recommended Negative:
> (worst quality, low quality:1.4), deformed, bad anatomy, bad hands, extra limbs, blurry, plastic skin, oversaturated, cartoon, 3d render
---
### โจ Key Features
- Pure Disentangled Training: Background, clothing, and hairstyles are trained with descriptive captions, allowing you to change outfits and locations freely without breaking the face.
- Natural Skin & Eyes: Preserves realistic skin pores, subtle highlights, and expressive gaze without the artificial "doll-like" plastic look.
- Universal SDXL Compatibility: Works seamlessly with all major realistic SDXL checkpoints and lighting styles.