Update app.py

app.py

@@ -102,7 +102,7 @@ def apply_llama3_chat_template(conversation, add_generation_prompt=True):
         prompt += "<|ASSISTANT|>\n"
     return prompt
 
-@spaces.GPU(duration=120)
+@spaces.GPU(duration=360)
 def generate_response(message, chat_history, system_prompt, temperature, max_tokens, top_p, top_k, repetition_penalty):
     # Build the conversation history.
     conversation = [{"role": "system", "content": system_prompt}]
@@ -163,10 +163,10 @@ with gr.Blocks(css=CSS, theme=gr.themes.Soft()) as demo:
 
     with gr.Accordion("⚙️ Settings", open=False):
         system_prompt = gr.TextArea(value=DEFAULT_SYSTEM_PROMPT, label="System Instructions")
-        temperature = gr.Slider(0, 1, value=0.3, label="Creativity (Temperature)")
+        temperature = gr.Slider(0, 1, value=0.6, label="Creativity (Temperature)")
         max_tokens = gr.Slider(128, 32768, 32768, label="Max Response Length")
         top_p = gr.Slider(0.0, 1.0, value=0.95, label="Top P (Nucleus Sampling)")
-        top_k = gr.Slider(0, 100, value=45, label="Top K")
+        top_k = gr.Slider(0, 100, value=35, label="Top K")
         repetition_penalty = gr.Slider(0.5, 2.0, value=1.1, label="Repetition Penalty")
 
     clear = gr.Button("Clear History")