Update app.py

app.py

@@ -36,23 +36,25 @@ def app(text, model):
 interface = gr.Interface(
     fn=app,
     inputs=[
-        gr.Textbox(label="Input Text", placeholder="Enter a sentence to visualize attention."),
+        gr.Textbox(label="Input Text", placeholder="Enter a sentence"),
         gr.Dropdown(
-            label="Choose a Transformer Model",
+            label="Model",
             choices=["bert-base-uncased", "distilbert-base-uncased", "roberta-base"],
             value=DEFAULT_MODEL
         )
     ],
     outputs=gr.Plot(label="Attention Map"),
-    title="🔍 Transformer Attention Visualizer",
-    description="Visualize how transformer models focus on different parts of input text using attention maps.",
-    article="""
-## How It Works
-This tool uses Hugging Face transformer models to extract **self-attention scores** from the last layer of the model.
-- 🧠 The attention map shows how each token attends to every other token.
-- 📊 You’re seeing Layer `-1` (last) and Head `0` for simplicity.
-- 🧪 Try different models and sentences to explore how they understand context differently.
-This is especially helpful for researchers, students, and NLP practitioners interested in interpretability.
+    title="Transformer Attention Visualizer",
+    description="""
+Understand how transformer models interpret text through self-attention.
+
+🧠 This tool extracts attention weights from the **last layer** and **first attention head** of popular transformer models.
+
+🔍 The attention map shows how each token focuses on others during processing.
+
+📚 Try different models and sentences to compare how they handle language and context.
+
+Ideal for NLP learners, researchers, and anyone curious about how transformers "pay attention".
 """
 )