Update app.py

app.py

@@ -14,6 +14,7 @@ local_rank, world_size = setup_model_parallel()
 generator = get_pretrained_models("7B", "tokenizer", local_rank, world_size)
 
 history = []
+simple_history = []
 
 def chat(user_input):
     bot_response = get_output(generator, user_input)[0]
@@ -26,16 +27,22 @@ def chat(user_input):
         "role": "system",
         "content": bot_response
     })    
+
+    simple_history.append((user_input, None))
     
     response = ""
     for word in bot_response.split(" "):
+        time.sleep(0.1)
         response += word + " "
-        yield [(user_input, response)]
+        current_pair = (user_input, response)
+        simple_history[-1] = current_pair
+        yield simple_history
 
 with gr.Blocks(css = """#col_container {width: 700px; margin-left: auto; margin-right: auto;}
                 #chatbot {height: 400px; overflow: auto;}""") as demo:
-    gr.Markdown(f"## {TITLE}\n\n\n\n{ABSTRACT}")
+    
     with gr.Column(elem_id='col_container'):
+        gr.Markdown(f"## {TITLE}\n\n\n\n{ABSTRACT}")
         chatbot = gr.Chatbot(elem_id='chatbot')
         textbox = gr.Textbox(placeholder="Enter a prompt")