add new

app.py

@@ -20,10 +20,6 @@ docs = []
 db = None
 extracted_text = None
 
-# ------------------------------
-# Title
-# ------------------------------
-st.title("📚 RAG For MSCI Indexes")
 
 # ------------------------------
 # Load Model for pretraining
@@ -47,35 +43,31 @@ def extract_text():
     documents = loader.load()
     return "\n".join([doc.page_content for doc in documents])
 
-
-with st.spinner("🔄 Loading Model..."):
-    generator = load_model()
-with st.spinner("🔄 Loading Knowldge Base..."):
-    extracted_text = extract_text()
-
-# ------------------------------
-# Extract Text
-# ------------------------------
-
-
-
-# ------------------------------
-# Build FAISS Index
-# ------------------------------
 @st.cache_resource
 def build_faiss(_docs):
     embeddings = HuggingFaceEmbeddings(model_name=embedding_model_name)
     return FAISS.from_documents(_docs, embeddings)
 
-if extracted_text:
+
+with st.spinner("🔄 Loading Model..."):
+    generator = load_model()
+with st.spinner("🔄 Loading Knowldge Base..."):
+    extracted_text = extract_text()
     splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)
     docs = [Document(page_content=chunk) for chunk in splitter.split_text(extracted_text)]
     db = build_faiss(docs)
     st.success("✅ Knowledge Base ready! From :- https://www.msci.com/indexes#featured-indexes")
 
+
+# ------------------------------
+# Title
+# ------------------------------
+st.title("📚 RAG For MSCI Indexes")
+st.markdown("This app uses a local LLM model to answer questions about MSCI Indexes using RAG (Retrieval Augmented Generation).")   
+
 query = st.text_input("💬 Ask a question about MSCI Indexes", placeholder="MSCI World IMI Index")
 
-if query and db:
+if query and db and extracted_text and len(docs) > 0:
     retriever = db.as_retriever(search_kwargs={"k": 3})
     retrieved_docs = retriever.get_relevant_documents(query)
     context = "\n".join([doc.page_content for doc in retrieved_docs])