Update app.py

app.py

@@ -8,14 +8,14 @@ import logging
 logging.basicConfig(level=logging.ERROR)
 logger = logging.getLogger(__name__)
 
-# Load a lightweight model for summarization (simulating LLM processing)
+# Use a lighter summarization model to reduce resource use
 try:
-    summarizer = pipeline("summarization", model="facebook/bart-large-cnn")
+    summarizer = pipeline("summarization", model="sshleifer/distilbart-cnn-6-6")  # Lighter than bart-large-cnn
 except Exception as e:
     logger.error(f"Summarizer loading failed: {e}")
     summarizer = None
 
-# Sample dataset of academic papers (title, abstract, content)
+# Sample dataset
 papers = [
     {
         "title": "Machine Learning in Healthcare",
@@ -34,14 +34,14 @@ papers = [
     }
 ]
 
-# Load pre-trained sentence transformer for embeddings
+# Load sentence transformer
 try:
     model = SentenceTransformer('all-MiniLM-L6-v2')
 except Exception as e:
     logger.error(f"SentenceTransformer loading failed: {e}")
     model = None
 
-# Pre-compute embeddings for the papers' content
+# Pre-compute embeddings
 paper_embeddings = model.encode([paper["content"] for paper in papers], convert_to_tensor=True) if model else None
 
 def semantic_search(query):
@@ -49,11 +49,11 @@ def semantic_search(query):
         return {"title": "Error", "abstract": "Error", "summary": "Model loading failed. Check logs.", "link": ""}
     
     try:
-        query_embedding = model.encode(query, convert_to_tensor=True)
+        query_embedding = model.encode([query], convert_to_tensor=True)  # Ensure list input
         similarities = np.dot(paper_embeddings, query_embedding.T).cpu().numpy()
         top_idx = np.argmax(similarities)
         top_paper = papers[top_idx]
-        summary = summarizer(top_paper["content"], max_length=50, min_length=25, do_sample=False)[0]["summary_text"]
+        summary = summarizer(top_paper["content"], max_length=30, min_length=10, do_sample=False)[0]["summary_text"]
         return {
             "title": top_paper["title"],
             "abstract": top_paper["abstract"],
@@ -67,7 +67,6 @@ def semantic_search(query):
 def search_interface(query):
     if not query:
         return "Please enter a search query.", "Error", "Error", "Error"
-    
     result = semantic_search(query)
     return (
         "✅ Search Complete!" if "Error" not in result["title"] else f"❌ Error: {result['summary']}",
@@ -76,22 +75,17 @@ def search_interface(query):
         result["summary"]
     )
 
-# Gradio UI
 with gr.Blocks(title="Semantic Search Engine for Academic Papers") as demo:
     gr.Markdown("# 🔍 Semantic Search Engine for Academic Papers\nSearch for academic papers by entering a research query.")
-    
     with gr.Row():
         query_input = gr.Textbox(label="📝 Enter Research Query", placeholder="e.g., 'machine learning in healthcare'")
         search_btn = gr.Button("🔎 Search")
-    
     with gr.Row():
         status = gr.Textbox(label="✅ Status", interactive=False)
         title_output = gr.Textbox(label="📑 Paper Title", interactive=False)
-    
     with gr.Row():
         abstract_output = gr.Textbox(label="📝 Abstract", interactive=False, lines=3)
         summary_output = gr.Textbox(label="📋 Summary", interactive=False, lines=3)
-
     search_btn.click(
         fn=search_interface,
         inputs=query_input,