Update README.md

# Argument Mining with LLaMA 3.1

This Space demonstrates a fine-tuned LLaMA 3.1 model for extracting arguments from historical newspapers.

## 🔍 Use Cases
- Identify argumentative passages in historical texts
- Support historians and linguists in discourse analysis
- Provide structured XML output for downstream research

## ⚡ Key Features
- Extracts **argumentation units**
- Provides structured `<argument>`, `<claim>`, `<explanation>` fields
- Multi-argument support per article
- Designed for **qualitative data research**

## 📊 Technical Details
- Base model: `Meta-Llama-3.1-8B-Instruct`
- Fine-tuned with LoRA adapters and 4-bit quantization
- Training data: German and English historical newspapers (disaster reportings)

---

README.md

@@ -3,9 +3,10 @@ title: Argument Mining
 colorFrom: blue
 colorTo: green
 sdk: gradio
-sdk_version: 5.25.0
+sdk_version: 5.44.1
 app_file: app.py
-pinned: false
+pinned: true
+licence: mit
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference