mainfile cgpt 5

app.py

@@ -67,7 +67,7 @@ def store_chat_history(role: str, content: str):
 def create_conversational_qa_chain(retriever, condense_llm, answer_llm):
     condense_question_chain = RunnableLambda(
         lambda x: {"chat_history": chistory, "question": x['question']}
-    ) | CONDENSE_QUESTION_PROMPT | condense_llm
+    ) | CONDENSE_QUESTION_PROMPT | RunnableLambda(lambda x: {"standalone_question": x})
 
     retrieval_chain = RunnableLambda(
         lambda x: x['standalone_question']
@@ -130,14 +130,18 @@ def app():
         conversational_qa_chain = create_conversational_qa_chain(retriever, llmc, llm)
         with st.chat_message("assistant"):
             with st.spinner("Thinking..."):
-                response = conversational_qa_chain.invoke(
-                {
-                    "question": prompts2,
-                    "chat_history": chistory,
-                }
-            )
-                st.write(response['generate_answer'])
-        message = {"role": "assistant", "content": response['generate_answer']}
+                response_chunks = []
+                async for chunk in conversational_qa_chain.astream(
+                    {
+                        "question": prompts2,
+                        "chat_history": chistory,
+                    }
+                ):
+                    response_chunks.append(chunk['generate_answer'])
+                    st.write("".join(response_chunks))
+                    
+        final_response = "".join(response_chunks)
+        message = {"role": "assistant", "content": final_response}
         st.session_state.messages.append(message)
 
 if __name__ == '__main__':