ffreemt commited on
Commit
bdaa897
1 Parent(s): e0faa34

Update paused

Browse files
Files changed (1) hide show
  1. app.py +6 -2
app.py CHANGED
@@ -351,6 +351,8 @@ with gr.Blocks(
351
  gr.Markdown(
352
  """<h4><center>mpt-30b-chat-ggml (q4_1)</center></h4>
353
 
 
 
354
  This demo is of [TheBloke/mpt-30B-chat-GGML](https://huggingface.co/TheBloke/mpt-30B-chat-GGML).
355
 
356
  Try to refresh the browser and try again when occasionally errors occur.
@@ -359,8 +361,8 @@ with gr.Blocks(
359
  """,
360
  elem_classes="xsmall",
361
  )
362
- conversation = Chat()
363
- chatbot = gr.Chatbot(scroll_to_output=True).style(height=700) # 500
364
  buff = gr.Textbox(show_label=False)
365
  with gr.Row():
366
  with gr.Column(scale=1):
@@ -440,6 +442,7 @@ with gr.Blocks(
440
  )
441
 
442
  _ = """
 
443
  submit_event = msg.submit(
444
  fn=conversation.user_turn,
445
  inputs=[msg, chatbot],
@@ -529,4 +532,5 @@ with gr.Blocks(
529
 
530
  # concurrency_count=5, max_size=20
531
  # max_size=36, concurrency_count=14
 
532
  block.queue(concurrency_count=5, max_size=20).launch(debug=True)
 
351
  gr.Markdown(
352
  """<h4><center>mpt-30b-chat-ggml (q4_1)</center></h4>
353
 
354
+ 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.
355
+
356
  This demo is of [TheBloke/mpt-30B-chat-GGML](https://huggingface.co/TheBloke/mpt-30B-chat-GGML).
357
 
358
  Try to refresh the browser and try again when occasionally errors occur.
 
361
  """,
362
  elem_classes="xsmall",
363
  )
364
+
365
+ chatbot = gr.Chatbot(value=[], scroll_to_output=True).style(height=700) # 500
366
  buff = gr.Textbox(show_label=False)
367
  with gr.Row():
368
  with gr.Column(scale=1):
 
442
  )
443
 
444
  _ = """
445
+ conversation = Chat()
446
  submit_event = msg.submit(
447
  fn=conversation.user_turn,
448
  inputs=[msg, chatbot],
 
532
 
533
  # concurrency_count=5, max_size=20
534
  # max_size=36, concurrency_count=14
535
+
536
  block.queue(concurrency_count=5, max_size=20).launch(debug=True)