[gradient_checkpointing] default to use it for torch 2.3 (#28538)

* default to use it

* style

src/transformers/modeling_utils.py

@@ -2092,7 +2092,7 @@ def gradient_checkpointing_enable(self, gradient_checkpointing_kwargs=None):
             raise ValueError(f"{self.__class__.__name__} does not support gradient checkpointing.")
 
         if gradient_checkpointing_kwargs is None:
-            gradient_checkpointing_kwargs = {}
+            gradient_checkpointing_kwargs = {"use_reentrant": True}
 
         gradient_checkpointing_func = functools.partial(checkpoint, **gradient_checkpointing_kwargs)