Update app.py

app.py

@@ -133,8 +133,8 @@ def generate_response(prompt):
     )
 
     device = "cuda" if torch.cuda.is_available() else "cpu"
-    
-    model = model.to(device)
+
+    model.to(device)
 
     
     inputs = {k: v.to(device) for k, v in inputs.items()}