Merge pull request #3982 from PaddlePaddle/fix_batch_norm_parameter_share

fix batch_norm parameter share

python/paddle/trainer/config_parser.py

@@ -2368,6 +2368,7 @@ class BatchNormLayer(LayerBase):
                  use_global_stats=True,
                  moving_average_fraction=0.9,
                  batch_norm_type=None,
+                 mean_var_names=None,
                  **xargs):
         if inputs is None:
             inputs = []
@@ -2421,6 +2422,11 @@ class BatchNormLayer(LayerBase):
 
         psize = self.calc_parameter_size(image_conf)
         dims = [1, psize]
+        if mean_var_names is not None:
+            assert len(mean_var_names) == 2
+            self.inputs[1].parameter_name = mean_var_names[0]
+            self.inputs[2].parameter_name = mean_var_names[1]
+
         self.create_input_parameter(0, psize)
         self.create_input_parameter(1, psize, dims)
         self.create_input_parameter(2, psize, dims)

python/paddle/trainer_config_helpers/layers.py

@@ -2959,7 +2959,8 @@ def batch_norm_layer(input,
                      layer_attr=None,
                      batch_norm_type=None,
                      moving_average_fraction=0.9,
-                     use_global_stats=None):
+                     use_global_stats=None,
+                     mean_var_names=None):
     """
     Batch Normalization Layer. The notation of this layer as follow.
 
@@ -3026,6 +3027,8 @@ def batch_norm_layer(input,
                                    :math:`runningMean = newMean*(1-factor)
                                    + runningMean*factor`
     :type moving_average_fraction: float.
+    :param mean_var_names: [mean name, variance name]
+    :type mean_var_names: string list
     :return: LayerOutput object.
     :rtype: LayerOutput
     """
@@ -3047,6 +3050,7 @@ def batch_norm_layer(input,
         bias=ParamAttr.to_bias(bias_attr),
         moving_average_fraction=moving_average_fraction,
         use_global_stats=use_global_stats,
+        mean_var_names=mean_var_names,
         **ExtraLayerAttribute.to_kwargs(layer_attr))
 
     return LayerOutput(