diff --git a/mindspore/train/model.py b/mindspore/train/model.py
index 75f1638a8fc..545b0330dcf 100755
--- a/mindspore/train/model.py
+++ b/mindspore/train/model.py
@@ -378,8 +378,8 @@ class Model:
         with _CallbackManager(callbacks) as list_callback:
             if not dataset_sink_mode:
                 self._train_process(epoch, train_dataset, list_callback, cb_params)
-            elif context.get_context("mode") == context.PYNATIVE_MODE:
-                logger.warning("The pynative mode cannot support dataset sink mode currently."
+            elif context.get_context("mode") == context.PYNATIVE_MODE or context.get_context("device_target") == "CPU":
+                logger.warning("The pynative mode and CPU cannot support dataset sink mode currently."
                                "So the training process will be performed with dataset not sink.")
                 self._train_process(epoch, train_dataset, list_callback, cb_params)
             else:
diff --git a/model_zoo/official/cv/lenet/train.py b/model_zoo/official/cv/lenet/train.py
index 7cd379134aa..1561b3daef1 100644
--- a/model_zoo/official/cv/lenet/train.py
+++ b/model_zoo/official/cv/lenet/train.py
@@ -44,8 +44,6 @@ if __name__ == "__main__":
 
     args = parser.parse_args()
 
-    if args.device_target == "CPU":
-        args.dataset_sink_mode = False
 
     context.set_context(mode=context.GRAPH_MODE, device_target=args.device_target)
     ds_train = create_dataset(os.path.join(args.data_path, "train"),