new updates

app.py

@@ -1,7 +1,7 @@
 import gradio as gr
 from transformers import pipeline
 
-# Load both summarization and sentiment models
+# Load models
 summarizer = pipeline(
     "summarization",
     model="Manish014/review-summariser-gpt-config1",
@@ -14,8 +14,8 @@ sentiment_analyzer = pipeline("sentiment-analysis")
 # Inference function
 def analyze_review(text):
     if not text.strip():
-        return "Please enter a review.", "Sentiment unavailable."
-
+        return "❗ Please enter a product review.", "❗ Sentiment unavailable."
+    
     summary = summarizer(
         text,
         max_length=80,
@@ -26,41 +26,50 @@ def analyze_review(text):
     )[0]["summary_text"]
 
     sentiment = sentiment_analyzer(text)[0]
-    sentiment_output = f"{sentiment['label']} ({round(sentiment['score'] * 100, 2)}%)"
+    sentiment_label = f"{sentiment['label']} ({round(sentiment['score'] * 100, 2)}%)"
 
-    return summary, sentiment_output
+    return summary, sentiment_label
 
-# Example reviews
-example_reviews = [
+# Example inputs
+examples = [
     ["This product leaks water and smells like burnt plastic."],
     ["Absolutely loved the screen resolution and battery life."],
-    ["Worst purchase I've made. Do not recommend at all."]
+    ["Worst purchase I've made. Do not recommend at all."],
+    ["The headphones are okay. Battery is good but fit is not comfortable."],
+    ["The fan is extremely loud and doesn't cool much."]
 ]
 
-# Build Gradio UI
+# Build UI
 with gr.Blocks(theme=gr.themes.Base()) as demo:
     gr.Markdown("## 📝 Review Summariser GPT - Config 1")
-    gr.Markdown("Enter a long product review below to get a helpful short summary and sentiment prediction.")
+    gr.Markdown("Enter a detailed product review below to receive a helpful summary ✂️ and predicted sentiment 📊.")
+
+    with gr.Row():
+        review_input = gr.Textbox(label="🗣️ Product Review", lines=5, placeholder="Write your review here...")
 
     with gr.Row():
-        input_box = gr.Textbox(label="🗣️ Product Review", lines=5, placeholder="Write your detailed review here...")
+        summary_output = gr.Textbox(label="✂️ Summary", lines=2)
+        sentiment_output = gr.Textbox(label="📊 Sentiment", lines=1)
 
     with gr.Row():
-        summary_output = gr.Textbox(label="✂️ Summary")
-        sentiment_output = gr.Textbox(label="📊 Sentiment")
+        analyze_btn = gr.Button("🔍 Analyze")
+        clear_btn = gr.Button("🧹 Clear")
 
-    submit_btn = gr.Button("🔍 Analyze Review")
-    clear_btn = gr.Button("🧹 Clear")
+    analyze_btn.click(analyze_review, inputs=review_input, outputs=[summary_output, sentiment_output])
+    clear_btn.click(lambda: ("", "", ""), outputs=[review_input, summary_output, sentiment_output])
 
-    submit_btn.click(analyze_review, inputs=input_box, outputs=[summary_output, sentiment_output])
-    clear_btn.click(lambda: ("", "", ""), outputs=[input_box, summary_output, sentiment_output])
+    gr.Examples(examples=examples, inputs=review_input, label="🔁 Try Example Reviews")
 
-    gr.Examples(
-        examples=example_reviews,
-        inputs=input_box,
-        label="🔁 Try an Example"
-    )
+    with gr.Accordion("ℹ️ About this App", open=False):
+        gr.Markdown(
+            """
+            This application uses a fine-tuned T5 model to summarize lengthy product reviews into short summaries and also classifies the sentiment as Positive or Negative.
+            - Model: `Manish014/review-summariser-gpt-config1`
+            - Summarization by 🤗 Transformers
+            - Sentiment by `distilbert-base-uncased-finetuned-sst-2-english`
+            """
+        )
 
-# Run the app
+# Run app
 if __name__ == "__main__":
     demo.launch()