🔮 Transform Your Images with Natural Language Instructions
Used with FluxFill-Dev The perfect companion for seamless editing
📝 Overview: The Future of Image Editing is Here
ICEdit-MoE-LoRA represents a breakthrough in AI image editing - a lightweight Low-Rank Adaptation (LoRA) module that brings professional-grade, instruction-based editing capabilities to large diffusion transformers. By intelligently fine-tuning only 1% of parameters using just 0.5% of the original training data, it achieves extraordinary editing capabilities that match or even surpass premium commercial solutions.
🛠️ Complete Installation Guide
📁 model Structure
📂 ComfyUI/
├── 📂 models/
│ ├── 📂 text_encoders/
│ │ ├── 📄 clip_l.safetensors
│ │ ├── 📄 t5xxl_fp16.safetensors
│ │ ├── 📄 t5xxl_fp8_e4m3fn.safetensors
│ │ └── 📄 t5xxl_fp8_e4m3fn_scaled.safetensors
│ ├── 📂 vae/
│ │ └── 📄 ae.safetensors
│ ├── 📂 diffusion_models/
│ │ ├── 📄 fluxfill-dev.safetensors
│ └── 📂 lora/
│ └── 📄 [ICEdit-MoE.safetensors💎 Essential Components
🔤 Text Encoders - The Brain Behind Natural Language Understanding
T5XXL Options - Choose based on your hardware capabilities:
🎭 VAE - The Visual Artist
🖼️ Diffusion Models - The Imaging Engine
✨ LoRA Models - The Magic Touch
🎬 Example Instructions You Can Try
"Add a majestic castle on the mountain in the background"
"Change the color of the car from red to electric blue"
"Remove the person from the image and replace with a flowering tree"
"Transform the style to watercolor painting while maintaining all details"
"Add dramatic sunset lighting to the scene"
🙏 Community Recognition
A special thank you to sanaka87 for revolutionizing open-source image editing by releasing the ICEdit-MoE-LoRA checkpoint and demos. Your contribution empowers creators everywhere and accelerates innovation across the AI art community!
My Node OLLAMAGEMINI
https://github.com/al-swaiti/ComfyUI-OllamaGemini
