diff --git a/python/paddle/distributed/fleet/data_generator/data_generator.py b/python/paddle/distributed/fleet/data_generator/data_generator.py
index 6fa6652e2c6bab3f29041c0ace5d701ab87d3b3f..cddba2ddee38219056d14eecc749a3d7eeac30eb 100644
--- a/python/paddle/distributed/fleet/data_generator/data_generator.py
+++ b/python/paddle/distributed/fleet/data_generator/data_generator.py
@@ -38,21 +38,20 @@ class DataGenerator:
 
             .. code-block:: python
 
-                import paddle.distributed.fleet.data_generator as dg
-                class MyData(dg.DataGenerator):
-
-                    def generate_sample(self, line):
-                        def local_iter():
-                            int_words = [int(x) for x in line.split()]
-                            yield ("words", int_words)
-                        return local_iter
-
-                    def generate_batch(self, samples):
-                        def local_iter():
-                            for s in samples:
-                                yield ("words", s[1].extend([s[1][0]]))
-                mydata = MyData()
-                mydata.set_batch(128)
+                >>> import paddle.distributed.fleet.data_generator as dg
+                >>> class MyData(dg.DataGenerator):
+                ...     def generate_sample(self, line):
+                ...         def local_iter():
+                ...             int_words = [int(x) for x in line.split()]
+                ...             yield ("words", int_words)
+                ...         return local_iter
+                ...
+                ...     def generate_batch(self, samples):
+                ...         def local_iter():
+                ...             for s in samples:
+                ...                 yield ("words", s[1].extend([s[1][0]]))
+                >>> mydata = MyData()
+                >>> mydata.set_batch(128)
 
         '''
         self.batch_size_ = batch_size
@@ -65,16 +64,15 @@ class DataGenerator:
         Example:
             .. code-block:: python
 
-                import paddle.distributed.fleet.data_generator as dg
-                class MyData(dg.DataGenerator):
-
-                    def generate_sample(self, line):
-                        def local_iter():
-                            yield ("words", [1, 2, 3, 4])
-                        return local_iter
-
-                mydata = MyData()
-                mydata.run_from_memory()
+                >>> # doctest: +SKIP('raise NotImplementedError')
+                >>> import paddle.distributed.fleet.data_generator as dg
+                >>> class MyData(dg.DataGenerator):
+                ...     def generate_sample(self, line):
+                ...         def local_iter():
+                ...             yield ("words", [1, 2, 3, 4])
+                ...         return local_iter
+                >>> mydata = MyData()
+                >>> mydata.run_from_memory()
         '''
         batch_samples = []
         line_iter = self.generate_sample(None)
@@ -104,17 +102,15 @@ class DataGenerator:
 
             .. code-block:: python
 
-                import paddle.distributed.fleet.data_generator as dg
-                class MyData(dg.DataGenerator):
-
-                    def generate_sample(self, line):
-                        def local_iter():
-                            int_words = [int(x) for x in line.split()]
-                            yield ("words", [int_words])
-                        return local_iter
-
-                mydata = MyData()
-                mydata.run_from_stdin()
+                >>> import paddle.distributed.fleet.data_generator as dg
+                >>> class MyData(dg.DataGenerator):
+                ...     def generate_sample(self, line):
+                ...         def local_iter():
+                ...             int_words = [int(x) for x in line.split()]
+                ...             yield ("words", [int_words])
+                ...         return local_iter
+                >>> mydata = MyData()
+                >>> mydata.run_from_stdin()
 
         '''
         batch_samples = []
@@ -177,15 +173,13 @@ class DataGenerator:
 
             .. code-block:: python
 
-                import paddle.distributed.fleet.data_generator as dg
-                class MyData(dg.DataGenerator):
-
-                    def generate_sample(self, line):
-                        def local_iter():
-                            int_words = [int(x) for x in line.split()]
-                            yield ("words", [int_words])
-                        return local_iter
-
+                >>> import paddle.distributed.fleet.data_generator as dg
+                >>> class MyData(dg.DataGenerator):
+                ...     def generate_sample(self, line):
+                ...         def local_iter():
+                ...             int_words = [int(x) for x in line.split()]
+                ...             yield ("words", [int_words])
+                ...         return local_iter
         '''
         raise NotImplementedError(
             "Please rewrite this function to return a list or tuple: "
@@ -210,21 +204,20 @@ class DataGenerator:
 
             .. code-block:: python
 
-                import paddle.distributed.fleet.data_generator as dg
-                class MyData(dg.DataGenerator):
-
-                    def generate_sample(self, line):
-                        def local_iter():
-                            int_words = [int(x) for x in line.split()]
-                            yield ("words", int_words)
-                        return local_iter
-
-                    def generate_batch(self, samples):
-                        def local_iter():
-                            for s in samples:
-                                yield ("words", s[1].extend([s[1][0]]))
-                mydata = MyData()
-                mydata.set_batch(128)
+                >>> import paddle.distributed.fleet.data_generator as dg
+                >>> class MyData(dg.DataGenerator):
+                ...     def generate_sample(self, line):
+                ...         def local_iter():
+                ...             int_words = [int(x) for x in line.split()]
+                ...             yield ("words", int_words)
+                ...         return local_iter
+                ...
+                ...     def generate_batch(self, samples):
+                ...         def local_iter():
+                ...             for s in samples:
+                ...                 yield ("words", s[1].extend([s[1][0]]))
+                >>> mydata = MyData()
+                >>> mydata.set_batch(128)
         '''
 
         def local_iter():