Update README.md

README.md

@@ -1,28 +1,37 @@
----
-license: mit
----
-# ModernBERT-CritiQ
-
-This model fine-tunes [ModernBERT-base](https://huggingface.co/answerdotai/ModernBERT-base) to classify critical questions as either:
+## 🔍 ModernBERT-CritiQ: Critical Question Usefulness Classifier
 
-- ✅ **Useful** (label = 1)
-- ❌ **Non-Useful** (label = 0)
+This model classifies **critical questions** as either `Useful` or `Non-Useful` in the context of an argument (intervention). It is fine-tuned on Critical Question Generation datasets using a ModernBERT architecture with extended context.
 
-## 🧠 Use Case
+---
 
-Evaluating the **usefulness** of questions in argumentation contexts (e.g., political debates, persuasive writing, education).
+## 💡 How to Use
 
-## 🧪 Dataset
+### 🛠 Install Dependencies
 
-Finetuned on: [`MidhunKanadan/CritiQ_BERT`](https://huggingface.co/datasets/MidhunKanadan/CritiQ_BERT)
+```bash
+pip install transformers
+```
 
-## 📦 Example
+### 🚀 Run with HuggingFace `pipeline`
 
 ```python
 from transformers import pipeline
 
 classifier = pipeline("text-classification", model="MidhunKanadan/ModernBERT-CritiQ")
 
-result = classifier("Are there more effective ways to improve fairness without raising the minimum wage?")
-print(result)
-# → [{'label': 'Useful', 'score': 0.9982688426971436}]
+intervention = "Investing in public transport reduces carbon emissions and benefits everyone."
+question = "What about the costs of implementing this system?"
+
+text = f"Intervention: {intervention} [SEP] Critical Question: {question}"
+result = classifier(text)[0]
+
+print(f"Label: {result['label']}, Confidence: {result['score']:.4f}")
+```
+
+---
+
+## 🧠 Expected Output
+
+```
+Label: Useful, Confidence: 0.9539
+```