Update app.py

app.py

@@ -85,28 +85,26 @@ def retrieve_and_generate_app(query, top_k=3):
     for i in I[0]:
         sol_id = document_ids[i]
         # Find the full content of the retrieved SOL
-        # This relies on the 'documents' list being correctly loaded and matching by ID
-        # --- CHANGE 1: Use 'text' key instead of 'content' here ---
-        retrieved_content = next((doc["text"] for doc in documents if doc["id"] == sol_id), "Content not found.")
-        retrieved_docs.append({"id": sol_id, "content": retrieved_content}) # Keep 'content' here for consistency in retrieved_docs structure if you like
+        # --- CHANGE THIS LINE ---
+        # Original (incorrect): retrieved_content = next((doc["text"] for doc in documents if doc["id"] == sol_id), "Content not found.")
+        retrieved_content = next((doc["content"] for doc in documents if doc["id"] == sol_id), "Content not found.")
+        retrieved_docs.append({"id": sol_id, "content": retrieved_content})
 
-    # --- CHANGE 2: Use 'text' key here for building the context ---
-    context = "\n\n".join([f"SOL {doc['id']}: {doc['text']}" for doc in retrieved_docs])
+    # 3. Context Construction
+    # --- CHANGE THIS LINE ---
+    # Original (incorrect): context = "\n\n".join([f"SOL {doc['id']}: {doc['text']}" for doc in retrieved_docs])
+    context = "\n\n".join([f"SOL {doc['id']}: {doc['content']}" for doc in retrieved_docs])
 
-    # --- CHANGE 3: Complete the prompt string ---
+    # 4. LLM Generation
     prompt = f"""
 Given the following information about Virginia Standards of Learning (SOLs):
-
 {context}
-
 Based on this information, answer the following question:
 {query}
-
 If the question is about a specific SOL number, provide a direct explanation for that SOL.
 If asked for lesson plans, worksheets, or proofs, explain what the document generally entails and whether it provides such materials.
 Be concise and to the point.
 """
-    # --- Start of the print statements for debugging (keep these for now!) ---
     print(f"\n--- PROMPT SENT TO LLM ---\n{prompt}\n--------------------------\n")
 
     response = llm_pipeline(prompt, max_new_tokens=500, num_return_sequences=1, do_sample=True, temperature=0.7)
@@ -123,8 +121,6 @@ Be concise and to the point.
         answer = generated_text
 
     print(f"\n--- FINAL ANSWER ---\n{answer}\n--------------------\n")
-    # --- End of the print statements for debugging ---
-
     return answer
 
     # Create Gradio interface