Update README.md

README.md

@@ -8,17 +8,9 @@ pinned: false
 ---
 
 ---
-title: QIMMA — Arabic LLM Leaderboard
+title: QIMMA — A quality-first Arabic LLM Leaderboard that evaluates and compares the performance of Arabic Large Language Models.
 ---
 
-<div align="center">
-
-# 🌟 QIMMA
-### The Open Arabic LLM Leaderboard
-
-*Community-driven benchmarking for Arabic large language models*
-
-**[📊 Leaderboard](https://huggingface.co/spaces/qimma/leaderboard)** · **[📦 Datasets](https://huggingface.co/datasets/qimma/leaderboard-results)** · **[💬 Discussions](https://huggingface.co/qimma/discussions)**
 
 </div>
 
@@ -32,33 +24,11 @@ The leaderboard is powered by a fully automated GPU evaluation pipeline, generou
 
 ---
 
-## Benchmarks
-
-| Benchmark | Status | What it measures |
-|-----------|--------|-----------------|
-| **FannOrFlop** | ✅ Active | Arabic creative text generation, scored with BERTScore against curated references |
-| **LightEval tasks** | 🔜 Coming soon | Arabic understanding and reasoning |
-| **EvalPlus / Code** | 🔜 Coming soon | Arabic-aware code generation |
-
----
-
 ## Submit Your Model
 
-Submissions are open to everyone. To submit, add a JSON file to the [leaderboard-requests](https://huggingface.co/datasets/qimma/leaderboard-requests) dataset:
-
-```json
-{
-  "model": "your-org/your-model-name",
-  "precision": "bfloat16",
-  "model_type": "instruct",
-  "status": "submitted",
-  "submitted_time": "2026-01-01T00:00:00Z",
-  "attempt": 1,
-  "max_attempts": 3
-}
-```
-
-Set `model_type` to `base` for base models or `instruct` for chat/instruction-tuned models.
+Submissions are open to everyone. To submit a model through the leaderboard space [QIMMA leaderboard](https://huggingface.co/spaces/qimma/leaderboard).
+
+Set `model type` to `base` for base models or `instruct` for chat/instruction-tuned models.
 
 ---