Update README.md

README.md

@@ -1,46 +1,20 @@
+---
+title: Rye AI
+emoji: 🐨
+colorFrom: yellow
+colorTo: purple
+thumbnail: >-
+  https://cdn-uploads.huggingface.co/production/uploads/651e9c3d8645f13f18eb3af4/ToSGmNsxw8iWNUPu63Eaj.png
+---
 # 🥪 Rye AI
+Open-source Danish NLP models and tools, layered like a proper smørrebrød.
 
-Welcome to **Rye AI** — a slightly absurd collection of Danish AI models, tools, and experiments… layered like a proper smørrebrød.
+### 🔪 What’s on the menu?
+- **DaMorph** – Morphological segmentation for Danish. Because long compound words deserve proper slicing.
+- **DaMedSum** – Medical summarization models trained on LUMI HPC, turning complex clinical text into clear, concise outputs.
+- **Morphology-Aware Tokenizers** – Danish tokenizers built to preserve linguistic structure and boost downstream performance.
 
-We build things that are:
-- open 🧑‍🍳  
-- layered 🧠  
-- occasionally overengineered 😏  
+### 🇩🇰 Why Rye AI?
+We build open, layered, and occasionally overengineered Danish NLP. Expect rigorous experiments, slightly playful ideas, and models that _probably_ shouldn’t work… but do.
 
-## 🍞 What’s on the menu?
-
-### 🧠 DaMorph
-A collection of experimental models exploring **morphological segmentation for Danish NLP**.
-
-👉 Because Danish words are long… and deserve to be sliced properly.
-
-**Includes:**
-- DaMorph
-- DaMorph Tokenizers (yes, we slice at every layer)
-- DaMedSum
-
-### 🏥 DaMedSum
-Danish medical summarization models trained on **LUMI HPC**.
-
-👉 Turning long, complicated medical text into something (slightly) more digestible.
-
-**Includes:**
-- T5-large
-- large / base / small variants  
-
-### 🔪 Tokenizers (a.k.a. precision slicing)
-Because no good sandwich starts without proper slicing.
-
-- Morphological tokenizers for Danish  
-- Built to explore how structure impacts understanding  
-
-## 🤔 What to expect
-
-- Serious experiments 🤓  
-- Slightly cursed ideas 😈  
-- Danish NLP in all its glory 🇩🇰  
-- Things that *probably* shouldn’t work… but do  
-
-## ⚡ Slogan
-
-**Open source never tasted this good.**
+> ⚡ **Open source never tasted this good.**