diff --git a/doc/api/v1/data_provider/pydataprovider2_en.rst b/doc/api/v1/data_provider/pydataprovider2_en.rst
index 30357be32538db4423ad0eaf899138256c84edc7..e8fb6292779790765154502bff319ea10ab1e70b 100644
--- a/doc/api/v1/data_provider/pydataprovider2_en.rst
+++ b/doc/api/v1/data_provider/pydataprovider2_en.rst
@@ -178,7 +178,7 @@ input_types
 +++++++++++
 
 PaddlePaddle has four data types, and three sequence types.
-The four data types are: 
+The four data types are:
 
 * :code:`dense_vector`: dense float vector.
 * :code:`sparse_binary_vector`: sparse binary vector, most of the value is 0, and
@@ -231,7 +231,7 @@ Its parameters lists as follows:
     * :code:`is_train` is a bool parameter that indicates the DataProvider is used in
       training or testing.
     * :code:`file_list` is the list of all files.
-      
+
   * User-defined parameters args can be set in training configuration.
 
 Note, PaddlePaddle reserves the right to add pre-defined parameter, so please