Update app.py

app.py

@@ -121,7 +121,8 @@ with gr.Blocks(css=CSS) as demo:
     with gr.Row():
         with gr.Column():
             gr.Markdown(f"""
-                    ## This demo is an unquantized GPU chatbot of [WizardCoder-Python-34B-V1.0-GGUF](https://huggingface.co/TheBloke/WizardCoder-Python-34B-V1.0-GGUF)
+                    ## This demo is a quantized GPU chatbot of [WizardCoder-Python-34B-V1.0-GGUF](https://huggingface.co/TheBloke/WizardCoder-Python-34B-V1.0-GGUF)
+                    It runs two different quantization levels in parallel for comparison. Best run at temperature 0.
                     """)
     with gr.Row():
         gr.Markdown("# 🔍 WizardCoder-Python-34B-V1.0-GGUF Playground Space! 🔎")
@@ -145,9 +146,9 @@ with gr.Blocks(css=CSS) as demo:
         with gr.Row():
             with gr.Column():
                 max_tokens = gr.Slider(20, 4000, label="Max Tokens", step=20, value=2000)
-                temperature = gr.Slider(0.0, 2.0, label="Temperature", step=0.1, value=0.8)
-                top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.02, value=0.95)
-                top_k = gr.Slider(-1, 100, label="Top K", step=1, value=40)
+                temperature = gr.Slider(0.0, 2.0, label="Temperature", step=0.1, value=0.0)
+                top_p = gr.Slider(0.0, 1.0, label="Top P", step=0.02, value=1.0)
+                top_k = gr.Slider(-1, 100, label="Top K", step=1, value=0)
                 repetition_penalty = gr.Slider(0.0, 2.0, label="Repetition Penalty", step=0.05, value=1.1)
 
         system_msg = gr.Textbox(