Nova

μ„ΈλΆ€ 정보

파일 λ‹€μš΄λ‘œλ“œ (1)

λͺ¨λΈ μ„€λͺ…

The following description is written by a neural network, because I am not a master of text (and a bit lazy)πŸ˜‚

### 🌟 Gorgeous Heterochromia & Freckles Character LoRA (SFW/NSFW)

This is a highly versatile and detailed character LoRA designed to capture a unique, high-quality female aesthetic. Fully trained from scratch with a balanced dataset to support both safe-for-work and explicit generations.

### πŸ”‘ Key Features Embedded in the Model:

You DO NOT need to type these tags to get them. They are baked directly into the character's core identity:

* Hair: Voluminous, beautiful curly red/auburn hair with bangs.

* Eyes: Stunning heterochromia (distinct multicolored eyes).

* Skin: Cute, detailed freckles covering the face and shoulders.

---

### πŸš€ How to Use & Prompting Guide:

* Trigger Word: nova (Always put it at the very beginning of your prompt).

#### πŸ›‘οΈ For SFW Generations:

The model has been trained to separate the character from her clothes. To change outfits, just prompt them normally:

> Example: nova, solo, upper body, looking at viewer, casual clothes, green hoodie, jeans, sitting in a cafe

#### πŸ”ž For NSFW / Explicit Content:

Thanks to a perfectly balanced 50/50 dataset, the model understands anatomy seamlessly without blending clothes into the skin:

> Example: nova, solo, full body, naked, nude, lying on bed, soft lighting

---

### βš™οΈ Recommended Settings:

* LoRA Weight: 0.8 - 1.0 (Start at 0.9 for the best balance of flexibility and character likeness).

* Base Models: Works perfectly with modern SDXL anime/semi-realistic checkpoints (Pony-based models, NoobAI, Animagine XL, etc.).

Feel free to share your generations in the review section! Enjoy!

이 λͺ¨λΈλ‘œ λ§Œλ“  이미지