Update README.md with selective FP8 quantization details and licensing

README.md

@@ -1,5 +1,66 @@
 ---
 license: other
-license_name: krea2
+license_name: krea-2-license
 license_link: https://www.krea.ai/krea-2-licensing
+tags:
+- text-to-image
+- image-generation
+- dit
+- fp8
+- comfyui
+- krea
+- krea2
 ---
+
+# Krea 2 OSS - Optimized FP8 Weights (Turbo)
+
+This repository provides an optimized **FP8 (float8_e4m3fn) weight-only quantized version** of the newly released **Krea 2 OSS (Turbo)** transformer. 
+
+This optimization reduces the model size from the original **24.76 GiB (BF16)** down to **12.01 GiB**, making it highly accessible and runnable on standard consumer hardware (such as 16GB and 24GB GPUs) without sacrificing output quality.
+
+## ⚠️ Licensing & Disclaimer
+- **Original Model Creators**: All credit goes to [KREA.ai](https://www.krea.ai) for the original research, architecture, and weights.
+- **License**: This model is subject to the **KREA 2 License Agreement**. Please read and comply with the official license terms before using these weights: [KREA 2 Licensing Terms](https://www.krea.ai/krea-2-licensing).
+- **Purpose**: This repository is a community-contributed utility. It does not claim ownership of the original model or architecture. Its sole purpose is to provide optimized, consumer-hardware-friendly weights for the open-source community.
+
+---
+
+## 🛠️ Quantization Details (Quality-First FP8)
+
+Unlike generic global quantization scripts that aggressively convert every parameter (which often degrades generation details or introduces NaN/promotion calculation errors in neural networks), this model was quantized using a **selective weight-only strategy**:
+
+1. **Targeted Quantization**: Only 2D floating-point weight matrices (`.weight` keys with `ndim >= 2` and element count `> 1024`) were quantized to `torch.float8_e4m3fn`.
+2. **Preserved Precision**: 
+   - All 1D vectors, biases, and normalization scales are kept in their native high-precision (`float32` / `bfloat16`).
+   - Highly sensitive projection/modulation layers (such as `LastLayer.modulation.lin` vectors) are **completely preserved** in high-precision. This prevents typical mathematical promotion bugs (such as `BFloat16` and `Float8` promotion issues in PyTorch) and retains original output fidelity.
+3. **Weight Comparison**:
+   - **Tensors Quantized to FP8**: 266 tensors.
+   - **Tensors Kept in Native Precision**: 166 tensors.
+   - **Size Reduction**: **24.76 GiB ➔ 12.01 GiB** (~51.5% VRAM / disk savings!).
+
+---
+
+## 🚀 How to Use in ComfyUI
+
+These weights are fully compatible with custom native ComfyUI nodes designed for the Krea2 pipeline.
+
+### 1. Model Placement
+Download the `krea2_turbo_fp8.safetensors` file from this repository and place it in your ComfyUI models folder:
+- Path: `ComfyUI/models/unet/` OR `ComfyUI/models/diffusion_models/` OR `ComfyUI/models/krea2/`
+
+### 2. Pipeline Dependencies
+To run Krea 2 OSS, you will also need the original text encoder and VAE:
+- **Text Encoder**: Qwen3-VL-4B-Instruct (`Qwen/Qwen3-VL-4B-Instruct` on Hugging Face).
+- **VAE**: Qwen-Image Autoencoder (`Qwen/Qwen-Image` subfolder `vae`).
+
+### 3. Recommended Generation Parameters (Turbo)
+- **Resolution**: 1024x1024 (up to 2048x2048)
+- **Steps**: 8
+- **CFG Scale**: 0.0 (Turbo model works best with 0 CFG)
+- **Mu**: 1.15
+- **Sampler / Scheduler**: Standard Flow Matching Timesteps
+
+---
+
+## 🤝 Acknowledgements
+Special thanks to the **KREA.ai** team for releasing Krea 2 to the open-source community. For any commercial licensing inquiries or details about the model, please visit [krea.ai](https://www.krea.ai).