Update app.py

app.py

@@ -1,3 +1,4 @@
+import spaces
 from threading import Thread
 
 import torch
@@ -20,6 +21,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_id)
 assistant_model = AutoModelForCausalLM.from_pretrained(assistant_id).to(torch_device)
 
 
+@spaces.GPU
 def run_generation(user_text, use_assistant, temperature, max_new_tokens):
     if temperature < 0.1:
         do_sample = False