Update README.md

README.md

@@ -1,3 +1,37 @@
----
-license: apache-2.0
----
+---
+license: apache-2.0
+base_model:
+- lodestones/Chroma
+---
+
+# Chroma FP8 Scaled
+
+## Model Details
+
+- **Model Type**: Scaled FP8 safetensors variant of Lodestone Rock's [Chroma](https://huggingface.co/lodestones/Chroma) model
+- **Model Architecture**: Chroma architecture, with FP8 scaling
+
+## Model Description
+
+Chroma FP8 Scaled is a high-precision variant of the Chroma model, utilizing the full dynamic range of FP8 (-448 to 448). This model leverages the large headroom available in FP8 format to maintain higher precision compared to standard FP8 safetensors, resulting in improved performance while maintaining the benefits of reduced model size.
+
+## Hardware and Software Requirements
+
+- **Dependencies**: Requires the `testing` branch of [FluxMod](https://github.com/lodestone-rock/ComfyUI_FluxMod)
+
+## Installation and Usage
+
+```
+# Clone the FluxMod repository
+git clone https://github.com/lodestone-rock/ComfyUI_FluxMod.git
+
+# Switch to the testing branch
+cd ComfyUI_FluxMod
+git checkout testing
+
+# Load the model using Chroma Unified Model Loader in ComfyUI
+```
+
+## Acknowledgments
+
+Thanks to Lodestone Rock for creating the original Chroma model and developing the FluxMod toolkit that enables this optimized FP8 representation.