Update paused

app.py

@@ -351,6 +351,8 @@ with gr.Blocks(
         gr.Markdown(
             """<h4><center>mpt-30b-chat-ggml (q4_1)</center></h4>
 
+            To run, a minimum of CPU UNGRADE hf instance is required. It takes around 60 seconds for initial response to appear. It can take a few minutes to complete a reply of decent length.
+
             This demo is of [TheBloke/mpt-30B-chat-GGML](https://huggingface.co/TheBloke/mpt-30B-chat-GGML).
 
             Try to refresh the browser and try again when occasionally errors occur.
@@ -359,8 +361,8 @@ with gr.Blocks(
             """,
             elem_classes="xsmall",
         )
-    conversation = Chat()
-    chatbot = gr.Chatbot(scroll_to_output=True).style(height=700)  # 500
+
+    chatbot = gr.Chatbot(value=[], scroll_to_output=True).style(height=700)  # 500
     buff = gr.Textbox(show_label=False)
     with gr.Row():
         with gr.Column(scale=1):
@@ -440,6 +442,7 @@ with gr.Blocks(
         )
 
     _ = """
+    conversation = Chat()
     submit_event = msg.submit(
         fn=conversation.user_turn,
         inputs=[msg, chatbot],
@@ -529,4 +532,5 @@ with gr.Blocks(
 
 # concurrency_count=5, max_size=20
 # max_size=36, concurrency_count=14
+
 block.queue(concurrency_count=5, max_size=20).launch(debug=True)