Disable @spaces.GPU decorator for T4 dedicated GPU

- Comment out @spaces.GPU() - not needed with dedicated T4
- Add comment explaining using T4 instead of ZeroGPU
- Model loads on startup and stays in memory
- No quota limits, always ready for production

app.py

@@ -68,7 +68,7 @@ def format_conversation_history(chat_history):
         messages.append({"role": role, "content": content})
     return messages
 
-@spaces.GPU()
+# @spaces.GPU()  # Disabled - using dedicated T4 GPU instead of ZeroGPU
 def generate_response(input_data, chat_history, max_new_tokens, temperature, top_p, top_k, repetition_penalty):
     new_message = {"role": "user", "content": input_data}
     processed_history = format_conversation_history(chat_history)