Update app.py

app.py

@@ -99,6 +99,7 @@ def load_new_model(model_id):
     except Exception as e:
         yield f"Error loading model: {str(e)}"
 
+@spaces.GPU
 def run_inference(user_prompt, max_tokens, temperature, top_k, top_p, rep_penalty, ngram_size, do_sample):
     """Generates text via streaming generator."""
     if model_manager.model is None or model_manager.tokenizer is None: