diff --git a/demo/mnist/api_train.py b/demo/mnist/api_train.py
index 77e6fa89003b6ff365e3fd0c98852ede0a4ab5ef..ae00e9f54c3a0e7da65c7738931ce3345bca98b2 100644
--- a/demo/mnist/api_train.py
+++ b/demo/mnist/api_train.py
@@ -71,8 +71,8 @@ def main():
     assert isinstance(updater, api.ParameterUpdater)
 
     # define network
-    images = paddle_v2.layer.data(name='pixel', type=dp.dense_vector(784))
-    label = paddle_v2.layer.data(name='label', type=dp.integer_value(10))
+    images = paddle_v2.layer.data(name='pixel', data_type=dp.dense_vector(784))
+    label = paddle_v2.layer.data(name='label', data_type=dp.integer_value(10))
     hidden1 = paddle_v2.layer.fc(input=images, size=200)
     hidden2 = paddle_v2.layer.fc(input=hidden1, size=200)
     inference = paddle_v2.layer.fc(input=hidden2,
diff --git a/demo/mnist/api_train_v2.py b/demo/mnist/api_train_v2.py
index 979cc3f1b9eb22d00cb49fe5ca2eab95dff78b56..b46cf1c8700ba8de2f9e5f8af3320ed6e1d9f7a9 100644
--- a/demo/mnist/api_train_v2.py
+++ b/demo/mnist/api_train_v2.py
@@ -16,8 +16,8 @@ def main():
     paddle.init(use_gpu=False, trainer_count=1)
 
     # define network topology
-    images = paddle.layer.data(name='pixel', type=dense_vector(784))
-    label = paddle.layer.data(name='label', type=integer_value(10))
+    images = paddle.layer.data(name='pixel', data_type=dense_vector(784))
+    label = paddle.layer.data(name='label', data_type=integer_value(10))
     hidden1 = paddle.layer.fc(input=images, size=200)
     hidden2 = paddle.layer.fc(input=hidden1, size=200)
     inference = paddle.layer.fc(input=hidden2,
diff --git a/python/paddle/v2/layer.py b/python/paddle/v2/layer.py
index e006b789222c0191d2e20506f1161323570e9b51..511b3e7457c261c2aed69c3a55ab34cd06bd4a78 100644
--- a/python/paddle/v2/layer.py
+++ b/python/paddle/v2/layer.py
@@ -165,26 +165,22 @@ So we also need to implement some special LayerV2.
 
 
 class DataLayerV2(Layer):
-    def __init__(self, name, type, **kwargs):
-        self.__method_name__ = 'data_layer'
-
-        assert isinstance(type, dp.InputType)
+    def __init__(self, name, data_type, **kwargs):
+        assert isinstance(data_type, dp.InputType)
 
-        # get data_size from type.dim
-        args = dict()
-        for key in kwargs:
-            args[key] = kwargs[key]
-        args['size'] = type.dim
-        self.__args__ = args
+        self.__method_name__ = 'data_layer'
+        self.__kwargs__ = kwargs
+        self.__data_size__ = data_type.dim
 
         super(DataLayerV2, self).__init__(name=name, parent_layers=dict())
 
     def to_proto_impl(self, **kwargs):
         args = dict()
+        args['size'] = self.__data_size__
         for each in kwargs:
             args[each] = kwargs[each]
-        for each in self.__args__:
-            args[each] = self.__args__[each]
+        for each in self.__kwargs__:
+            args[each] = self.__kwargs__[each]
         return getattr(conf_helps, self.__method_name__)(name=self.name, **args)
 
 
@@ -202,8 +198,8 @@ cross_entropy_cost = __convert_to_v2__(
     parent_names=['input', 'label'])
 
 if __name__ == '__main__':
-    pixel = data(name='pixel', type=dp.dense_vector(784))
-    label = data(name='label', type=dp.integer_value(10))
+    pixel = data(name='pixel', data_type=dp.dense_vector(784))
+    label = data(name='label', data_type=dp.integer_value(10))
     hidden = fc(input=pixel, size=100, act=conf_helps.SigmoidActivation())
     inference = fc(input=hidden, size=10, act=conf_helps.SoftmaxActivation())
     maxid = max_id(input=inference)