diff --git a/python/paddle/fluid/dataloader/dataloader_iter.py b/python/paddle/fluid/dataloader/dataloader_iter.py
index f55ea1d9637923b2cee6d221f8a8d856ad9dfe86..9756936f57990112a586ac7eee3324fb09462601 100644
--- a/python/paddle/fluid/dataloader/dataloader_iter.py
+++ b/python/paddle/fluid/dataloader/dataloader_iter.py
@@ -320,7 +320,6 @@ class _DataLoaderIterSingleProcess(_DataLoaderIterBase):
                 array = core.LoDTensorArray()
                 for slot in batch:
                     if not isinstance(slot, core.LoDTensor):
-                        self._check_input_array(slot)
                         # FIXME(dkp): blocking_queue only support
                         #             core.LoDTensorArray as input now, read
                         #             numpy data into a LoDTensorArray here,
@@ -346,19 +345,6 @@ class _DataLoaderIterSingleProcess(_DataLoaderIterBase):
             logging.warning("DataLoader reader thread raised an exception.")
             six.reraise(*sys.exc_info())
 
-    @classmethod
-    def _check_input_array(cls, item):
-        if isinstance(item, paddle.Tensor):
-            return
-        arr = np.array(item)
-        if arr.dtype == np.object:
-            raise TypeError((
-                "\n\tFaild to convert input data to a regular ndarray :\n\t* Usually "
-                "this means the input data contains nested lists with different lengths. "
-                "\n\t* Check the reader function passed to 'decorate_batch_generator'"
-                " to locate the data causes this issue.\n\t* Please consider using "
-                "'fluid.create_lod_tensor' to convert it to a LoD-Tensor."))
-
     def __next__(self):
         try:
             if in_dygraph_mode():