diff --git a/python/paddle/v2/parameters.py b/python/paddle/v2/parameters.py
index a9cba8ca0b1ef..f994602a6255d 100644
--- a/python/paddle/v2/parameters.py
+++ b/python/paddle/v2/parameters.py
@@ -113,16 +113,7 @@ def __iter__(self):
         """
         return iter(self.__param_conf__)
 
-    def __getitem__(self, key):
-        """
-        Get parameter by parameter name. It uses Python dict syntax.
-
-        :note: It will always copy the parameter from C++ side.
-        :param key: Parameter name
-        :type key: basestring
-        :return: parameter value
-        :rtype: np.ndarray
-        """
+    def __getter_inner(self, key, param_type):
         import py_paddle.swig_paddle as api
         shape = self.get_shape(key)
 
@@ -138,7 +129,7 @@ def __getitem__(self, key):
                     each_gradient_machine, key)
                 # for simplify implementation now, we always copy from C++
                 assert isinstance(param, api.Parameter)
-                val = param.getBuf(api.PARAMETER_VALUE)
+                val = param.getBuf(param_type)
                 assert isinstance(val, api.Vector)
                 val = val.copyToNumpyArray()
                 return val
@@ -146,6 +137,19 @@ def __getitem__(self, key):
 
             raise RuntimeError("Unexpected branch")
 
+    def __getitem__(self, key):
+        """
+        Get parameter by parameter name. It uses Python dict syntax.
+
+        :note: It will always copy the parameter from C++ side.
+        :param key: Parameter name
+        :type key: basestring
+        :return: parameter value
+        :rtype: np.ndarray
+        """
+        import py_paddle.swig_paddle as api
+        return self.__getter_inner(key, api.PARAMETER_VALUE)
+
     def get_shape(self, key):
         """
         get shape of the parameter.
@@ -202,6 +206,19 @@ def get(self, parameter_name):
         """
         return self.__getitem__(key=parameter_name)
 
+    def get_grad(self, key):
+        """
+        Get grandient by parameter name.
+
+        :note: It will always copy the parameter from C++ side.
+        :param key: parameter name
+        :type key: basestring
+        :return: The grandient matrix.
+        :rtype: np.ndarray
+        """
+        import py_paddle.swig_paddle as api
+        return self.__getter_inner(key, api.PARAMETER_GRADIENT)
+
     def set(self, parameter_name, value):
         """
         Set parameter by parameter name & matrix.
diff --git a/python/paddle/v2/trainer.py b/python/paddle/v2/trainer.py
index 76bae0bb12b6c..9c4dd5f25083d 100644
--- a/python/paddle/v2/trainer.py
+++ b/python/paddle/v2/trainer.py
@@ -161,14 +161,14 @@ def train(self, reader, num_passes=1, event_handler=None, feeding=None):
                     self.__parameter_updater__.update(each_param)
                 cost_sum = out_args.sum()
                 cost = cost_sum / len(data_batch)
-                self.__parameter_updater__.finishBatch(cost)
-                batch_evaluator.finish()
                 event_handler(
                     v2_event.EndIteration(
                         pass_id=pass_id,
                         batch_id=batch_id,
                         cost=cost,
                         evaluator=batch_evaluator))
+                self.__parameter_updater__.finishBatch(cost)
+                batch_evaluator.finish()
 
             self.__parameter_updater__.finishPass()
             pass_evaluator.finish()