Update app.py

app.py

@@ -139,30 +139,33 @@ def conversation(qa_chain, message, history):
     # Append user message and response to chat history
     new_history = history + [(message, response_answer)]
     # return gr.update(value=""), new_history, response_sources[0], response_sources[1] 
-    return qa_chain, gr.update(value=""), new_history, response_source1, response_source1_page, response_source2, response_source2_page, response_source3, response_source3_page
+    return qa_chain, new_history, response_source1, response_source1_page, response_source2, response_source2_page, response_source3, response_source3_page
     
 #document = os.listdir(list_file_obj)
-vector_db, collection_name = initialize_database(list_file_obj)
 #qa_chain = 
 
-# Initialize langchain LLM chain
-llm = HuggingFaceHub(repo_id = llm_model,
-                    model_kwargs={"temperature": temperature, 
-                                  "max_new_tokens": max_tokens, 
-                                  "top_k": top_k, 
-                                  "load_in_8bit": True})
-retriever=vector_db.as_retriever()
-memory = ConversationBufferMemory(memory_key="chat_history", output_key='answer', return_messages=True)
-qa_chain = ConversationalRetrievalChain.from_llm(llm,retriever=retriever,chain_type="stuff", 
-                                                 memory=memory,return_source_documents=True,verbose=False,)
 
-print('qa chain and vector_db done')
+
 
 def demo():
     with gr.Blocks(theme='base') as demo:
-        #vector_db = gr.State()
-        #qa_chain = gr.State()
-        #collection_name = gr.State()
+        vector_db = gr.State()
+        qa_chain = gr.State()
+        collection_name = gr.State()
+
+
+        vector_db, collection_name = initialize_database(list_file_obj)
+
+        # Initialize langchain LLM chain
+        llm = HuggingFaceHub(repo_id = llm_model,
+                            model_kwargs={"temperature": temperature, 
+                                          "max_new_tokens": max_tokens, 
+                                          "top_k": top_k, 
+                                          "load_in_8bit": True})
+        retriever=vector_db.as_retriever()
+        memory = ConversationBufferMemory(memory_key="chat_history", output_key='answer', return_messages=True)
+        qa_chain = ConversationalRetrievalChain.from_llm(llm,retriever=retriever,chain_type="stuff", 
+                                                         memory=memory,return_source_documents=True,verbose=False,)
         
         chatbot = gr.Chatbot(height=300)
         with gr.Accordion('References', open=True):