Upload README.md with huggingface_hub

README.md

@@ -1,10 +1,30 @@
 ---
 title: README
-emoji: 🦀
-colorFrom: yellow
-colorTo: red
+emoji: 🧩
+colorFrom: purple
+colorTo: pink
 sdk: static
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# diffusers-modular
+
+Creative workflows built with **Modular Diffusers** — the demos, and the checkpoints they run on.
+
+## What Modular Diffusers is
+
+Diffusion pipelines are usually monolithic: one class per task, and a new task means a new pipeline. Modular
+Diffusers breaks that into **composable blocks** — encode, prepare, denoise, decode — that you assemble into a
+workflow, share components between them through a `ComponentsManager` so one loaded model serves several tasks,
+and swap or subclass a single block instead of forking a pipeline. It is how a workflow that would be a graph of
+custom nodes elsewhere becomes a few lines of Python.
+
+- 📚 [Documentation](https://huggingface.co/docs/diffusers/main/en/modular_diffusers/overview) · [Quickstart](https://huggingface.co/docs/diffusers/main/en/modular_diffusers/quickstart)
+- 📝 [Introducing Modular Diffusers](https://huggingface.co/blog/modular-diffusers)
+- 🧪 [Modular Diffusers on GitHub](https://github.com/huggingface/diffusers)
+
+## In this org
+
+- [**MiniMax-H3-Pruned-Ref-Delta-Fused-r1024**](https://huggingface.co/diffusers-modular/MiniMax-H3-Pruned-Ref-Delta-Fused-r1024)
+  — one MiniMax-H3 transformer that serves both reference conditioning *and* first/last-frame conditioning, where
+  the model normally needs two separate 37.5 GB partitions.