Update app.py

app.py

@@ -5,72 +5,82 @@ from langchain.text_splitter import RecursiveCharacterTextSplitter
 from langchain_community.vectorstores import Chroma
 from langchain_community.embeddings import HuggingFaceEmbeddings
 from langchain_groq import ChatGroq
-from langchain.document_loaders import PyPDFLoader
+from langchain_community.document_loaders import PyPDFLoader
 
-# --- Hugging Face Hub Setup ---
-HF_REPO_ID = "Shami96/7solar-documentation"  # Replace with your dataset
-HF_PDF_NAME = "7solar_documentation.pdf"  # Your PDF filename
+# Configuration
+HF_REPO_ID = "Shami96/7solar-documentation"
+HF_PDF_NAME = "7solar_documentation.pdf"
+HF_TOKEN = os.environ.get("HF_TOKEN")
 
-# --- Load PDF from Hugging Face Hub ---
-def load_pdf_from_hf():
-    pdf_path = hf_hub_download(
-        repo_id=HF_REPO_ID,
-        filename=HF_PDF_NAME,
-        repo_type="dataset",  # Critical for datasets!
-        token=os.environ.get("HF_TOKEN")  # For private repos
-    )
-    loader = PyPDFLoader(pdf_path)
-    return loader.load()
+# Initialize components
+def initialize_components():
+    print("⚙️ Initializing components...")
+    
+    # Load PDF
+    try:
+        pdf_path = hf_hub_download(
+            repo_id=HF_REPO_ID,
+            filename=HF_PDF_NAME,
+            repo_type="dataset",
+            token=HF_TOKEN
+        )
+        loader = PyPDFLoader(pdf_path)
+        documents = loader.load()
+    except Exception as e:
+        raise RuntimeError(f"Failed to load PDF: {str(e)}")
 
-# --- Split & Embed Docs ---
-def create_vector_db():
-    docs = load_pdf_from_hf()
+    # Create vector store
     text_splitter = RecursiveCharacterTextSplitter(
-        chunk_size=2000,
-        chunk_overlap=300
+        chunk_size=1000,
+        chunk_overlap=200
     )
-    chunks = text_splitter.split_documents(docs)
+    chunks = text_splitter.split_documents(documents)
     embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
     return Chroma.from_documents(chunks, embeddings)
 
-# --- RAG Chatbot Logic ---
-def get_response(query, history):
-    # Handle greetings
-    if query.lower() in ["hi", "hello", "hey"]:
-        return history + [(query, "Hello! 👋 Ask me about 7Solar's solar packages or services!")]
-
-    # Retrieve relevant doc chunks
-    matching_docs = vector_db.similarity_search(query, k=5)
-    if not matching_docs:
-        return history + [(query, "I couldn't find details. Ask about 7Solar's services!")]
-
-    # Generate LLM response
-    llm = ChatGroq(
-        model_name="llama3-70b-8192",
-        temperature=0.2,
-        api_key=os.environ.get("GROQ_API_KEY")  # Set in Spaces Secrets
-    )
-    context = "\n\n".join([doc.page_content for doc in matching_docs])
-    response = llm.invoke(
-        f"Answer this query using ONLY the text below:\n\n{context}\n\nQuestion: {query}"
-    )
-    return history + [(query, response.content)]
-
-# --- Initialize Vector DB ---
-print("⚙️ Loading document...")
-vector_db = create_vector_db()
+# Chat function
+def respond(message, history):
+    try:
+        # Initialize if not already done
+        if 'vector_db' not in globals():
+            global vector_db
+            vector_db = initialize_components()
+        
+        # Handle greetings
+        if message.lower() in ["hi", "hello", "hey"]:
+            return "Hello! I'm your 7Solar assistant. How can I help you today?"
+        
+        # Search documents
+        docs = vector_db.similarity_search(message, k=3)
+        if not docs:
+            return "I couldn't find relevant information. Please try another question about 7Solar."
+        
+        # Generate response
+        llm = ChatGroq(
+            model_name="llama3-70b-8192",
+            temperature=0.3
+        )
+        context = "\n\n".join([doc.page_content for doc in docs])
+        response = llm.invoke(
+            f"Using only this context:\n{context}\n\nQuestion: {message}\nAnswer:"
+        )
+        return response.content
+    except Exception as e:
+        return f"An error occurred: {str(e)}"
 
-# --- Gradio Interface ---
-with gr.Blocks() as demo:
-    gr.Markdown("# ☀️ 7Solar Smart Assistant")
-    chatbot = gr.Chatbot()
-    msg = gr.Textbox(label="Ask about solar packages, services, etc.")
-    msg.submit(get_response, [msg, chatbot], [chatbot])
-    clear = gr.Button("Clear Chat")
-    clear.click(lambda: [], None, chatbot, queue=False)
+# Create Gradio interface
+demo = gr.ChatInterface(
+    fn=respond,
+    title="☀️ 7Solar Assistant",
+    description="Ask me anything about 7Solar's services and documentation",
+    examples=["What solar packages do you offer?", "How does the registration process work?"],
+    cache_examples=False
+)
 
-demo.launch(
-    server_name="0.0.0.0",
-    server_port=7860,
-    #enable_api=True  # This explicitly enables the API
-)
+# Launch with error handling
+if __name__ == "__main__":
+    try:
+        demo.launch(server_name="0.0.0.0", server_port=7860)
+    except Exception as e:
+        print(f"Failed to launch: {str(e)}")
+        raise