Update app.py

app.py

@@ -24,18 +24,8 @@ client = None
 def load_llm():
     # Ensure you have the API Key set in your environment or via input
     api_key = os.getenv("CEREBRAS_API_KEY")
-    if not api_key:
-        api_key = st.text_input("Enter your Cerebras API key:")
-    if api_key:
-        llm = Cerebras(model="llama-3.3-70b", api_key=api_key)
-        return llm
-    else:
-        st.error("API Key is required.")
-        return None
-
-# Load llm at the beginning of the session
-if "llm" not in st.session_state:
-    st.session_state.llm = load_llm()
+    llm = Cerebras(model="llama-3.3-70b", api_key=api_key)
+    return llm
 
 def reset_chat():
     st.session_state.messages = []
@@ -46,8 +36,6 @@ def display_excel(file):
     st.markdown("### Excel Preview")
     # Read the Excel file
     df = pd.read_excel(file)
-    # Attempt to fix mixed-type column errors
-    #df = df.apply(pd.to_numeric, errors='coerce')  # Convert columns to numeric where possible
     # Display the dataframe
     st.dataframe(df)
 
@@ -149,23 +137,18 @@ if prompt := st.chat_input("What's up?"):
     with st.chat_message("assistant"):
         message_placeholder = st.empty()
         full_response = ""
+        
+        # Simulate stream of response with milliseconds delay
+        streaming_response = query_engine.query(prompt)
+        
+        for chunk in streaming_response.response_gen:
+            full_response += chunk
+            message_placeholder.markdown(full_response + "▌")
 
-        # Ensure llm is loaded
-        if st.session_state.llm:
-            # Using Cerebras stream_chat for streaming response
-            messages = [
-                ChatMessage(role="user", content=prompt)
-            ]
+        # full_response = query_engine.query(prompt)
 
-            response = st.session_state.llm.stream_chat(messages)
-            for r in response:
-                full_response += r.delta
-                message_placeholder.markdown(full_response + "▌")
-
-            message_placeholder.markdown(full_response)
-
-        else:
-            st.error("LLM model is not initialized correctly.")
+        message_placeholder.markdown(full_response)
+        # st.session_state.context = ctx
 
     # Add assistant response to chat history
-    st.session_state.messages.append({"role": "assistant", "content": full_response})
+    st.session_state.messages.append({"role": "assistant", "content": full_response})