README.md

---
title: Kontext-dev LORA (Na'vi character transformation)
emoji: 🧞‍♂️
colorFrom: blue
colorTo: purple
sdk: gradio
sdk_version: "5.40.0"
app_file: app.py
pinned: false
---

# 🧞‍♂️ Avatar Transformation Studio

Transform any portrait into a stunning Na'vi character with bioluminescent blue skin and mystical Avatar features using FLUX Kontext + Avatar LoRA.

## ✨ Features

- **FLUX Kontext Model**: Advanced image-to-image transformation
- **Avatar LoRA**: Specialized fine-tuning for Na'vi character generation
- **Real-time GPU Processing**: Fast inference with Spaces GPU
- **Custom Prompts**: Add your own creative details to the transformation
- **High Quality Output**: Photorealistic Avatar character generation

## 🚀 How to Use

1. Upload a clear portrait photo (front-facing works best)
2. Optionally add custom details like "warrior markings", "glowing tattoos", etc.
3. Click "Transform to Na'vi Avatar" 
4. Wait for your magical transformation!

## 💡 Tips for Best Results

- 📷 Use clear, well-lit photos
- 👤 Front-facing portraits work best  
- 🎨 High resolution images recommended
- ✨ Be creative with custom prompts

## 🛠️ Technical Details

- **Base Model**: FLUX.1-Kontext-dev
- **LoRA**: Avataar_LoRA_000003000
- **Framework**: ComfyUI + Gradio
- **GPU**: Spaces GPU (T4/A10G)
- **Processing Time**: ~30-60 seconds per image

## 📋 Requirements

The space automatically downloads and manages all required models:
- FLUX Kontext checkpoint (~12GB)
- Avatar LoRA weights (~200MB)
- VAE and text encoders (~5GB)

## 🎯 Model Credits

- **FLUX Kontext**: [black-forest-labs/FLUX.1-Kontext-dev](https://huggingface.co/black-forest-labs/FLUX.1-Kontext-dev)
- **Avatar LoRA**: [DoozyWo/Kontext_avatar_LoRA](https://huggingface.co/DoozyWo/Kontext_avatar_LoRA)
- **Text Encoders**: [comfyanonymous/flux_text_encoders](https://huggingface.co/comfyanonymous/flux_text_encoders)

## ⚠️ Usage Notes

- First run may take longer due to model downloads
- GPU allocation limited to 60 seconds per generation
- Best results with human portraits
- Generated images are for creative/entertainment purposes

---

*Built with ❤️ using ComfyUI, FLUX, and Gradio*