Update app.py

app.py

@@ -69,6 +69,7 @@ def load_model(model_name):
         # Only move to CUDA if it's not a quantized model
         if model_name not in quantized_models:
             model = model.to("cuda")
+            tokenizer = tokenizer.to("cuda")
         
         selected_model = model_name
     except Exception as e: