Update app.py

app.py

@@ -1,5 +1,4 @@
 import gradio as gr
-import spaces
 from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer
 import torch
 from threading import Thread
@@ -11,7 +10,6 @@ device = "cuda:0" if torch.cuda.is_available() else "cpu"
 phi4_model = AutoModelForCausalLM.from_pretrained(phi4_model_path, device_map="auto", torch_dtype="auto")
 phi4_tokenizer = AutoTokenizer.from_pretrained(phi4_model_path)
 
-@spaces.GPU(duration=60)
 def generate_response(user_message, max_tokens, temperature, top_k, top_p, repetition_penalty, history_state):
     if not user_message.strip():
         return history_state, history_state