Update app.py

app.py

@@ -108,7 +108,7 @@ ci = gr.ChatInterface(
 
 with gr.Blocks(css=".bubble-gap {gap: 6px !important}", theme="shivi/calm_seafoam") as demo:
   gr.Markdown("# EliGPT v1.3")
-  gr.Markdown("Llama 3 8b finetuned on 2.5k of my discord messages. [Train your own clone!](https://gist.github.com/Elijah-Bodden/1964bd02fcd19efef65f6e0cd92881c4)\n(The model can be slow when multiple people are using it. Duplicate the space to get your own free faster instance)")
+  gr.Markdown("Llama 3 8b finetuned on 2.5k of my discord messages. [Train your own clone!](https://gist.github.com/Elijah-Bodden/1964bd02fcd19efef65f6e0cd92881c4)\n(The model can be slow when multiple people are using it. Duplicate the space to get your own free faster instance. It can also be slow to start up if it hasn't been run in a while.)")
   # with gr.Accordion("Q&A:", open=False):
   #     gr.Markdown("""Q: Why is the model so fucking slow  
   #       A: The model might be slow if it hasn't run recently or a lot of people are using it (it's running on llama.cpp on a single a very slow cpu). You can duplicate the space to get your own (free) instance with no wait times.