Fix distillation for soft label. (#16538)

test=develop

python/paddle/fluid/contrib/slim/distillation/distiller.py

@@ -19,7 +19,7 @@ from .... import Program
 from .... import program_guard
 from .... import regularizer
 
-__all__ = ['FSPDistiller', 'L2Distiller']
+__all__ = ['FSPDistiller', 'L2Distiller', 'SoftLabelDistiller']
 
 
 class L2Distiller(object):
@@ -186,3 +186,91 @@ class FSPDistillerPass(object):
 
     def _fsp_matrix(self, fea_map_0, fea_map_1):
         return layers.fsp_matrix(fea_map_0, fea_map_1)
+
+
+class SoftLabelDistiller(object):
+    """
+    Combine two layers from student net and teacher net by softmax_with_cross_entropy loss.
+    And add the loss into the total loss using for distillation training.
+    """
+
+    def __init__(self,
+                 student_feature_map=None,
+                 teacher_feature_map=None,
+                 student_temperature=1.0,
+                 teacher_temperature=1.0,
+                 distillation_loss_weight=1):
+        """
+        Args:
+            student_feature_map(str): The name of feature map from student network.
+            teacher_feature_map(str): The name of feature map from teacher network.
+                                      It's shape should be the same with student network.
+            student_temperature(float): Temperature used to divide student_feature_map before softmax_with_cross_entropy. default: 1.0
+            teacher_temperature(float): Temperature used to divide teacher_feature_map before softmax_with_cross_entropy. default: 1.0
+            distillation_loss_weight(float): The weight of the l2-loss.
+        """
+
+        self.student_feature_map = student_feature_map
+        self.teacher_feature_map = teacher_feature_map
+        self.distillation_loss_weight = distillation_loss_weight
+        self.student_temperature = student_temperature
+        self.teacher_temperature = teacher_temperature
+
+    def distiller_loss(self, graph):
+        """
+        Modify graph inplace to add softmax_with_cross_entropy loss.
+        Args: 
+            graph(GraphWrapper): The graph to be modified.
+        Returns:
+            GraphWrapper: The modified graph.
+        """
+        distiller_pass = SoftLabelDistillerPass(
+            self.student_feature_map, self.teacher_feature_map,
+            self.student_temperature, self.teacher_temperature,
+            self.distillation_loss_weight)
+        dis_graph = distiller_pass.apply(graph)
+        return dis_graph
+
+
+class SoftLabelDistillerPass(object):
+    def __init__(self,
+                 student_feature_map,
+                 teacher_feature_map,
+                 student_temperature,
+                 teacher_temperature,
+                 distillation_loss_weight=1):
+        """
+        Args:
+            student_feature_map(str): The name of feature map from student network.
+            teacher_feature_map(str): The name of feature map from teacher network.
+                                      It's shape should be the same with student network.
+            student_temperature(float): Temperature used to divide student_feature_map before softmax_with_cross_entropy.
+            teacher_temperature(float): Temperature used to divide teacher_feature_map before softmax_with_cross_entropy.
+            distillation_loss_weight(float): The weight of the l2-loss.
+        """
+        self.student_feature_map = student_feature_map
+        self.teacher_feature_map = teacher_feature_map
+        self.student_temperature = student_temperature
+        self.teacher_temperature = teacher_temperature
+        self.distillation_loss_weight = distillation_loss_weight
+
+    def apply(self, graph):
+        ret_graph = graph
+        with program_guard(ret_graph.program):
+
+            student_feature_map = ret_graph.var(self.student_feature_map)._var
+            teacher_feature_map = ret_graph.var(self.teacher_feature_map)._var
+            s_fea = student_feature_map / self.student_temperature
+            t_fea = teacher_feature_map / self.distillation_loss_weight
+            t_fea.stop_gradient = True
+            ce_loss = layers.softmax_with_cross_entropy(
+                s_fea, t_fea, soft_label=True)
+            distillation_loss = ce_loss * self.distillation_loss_weight
+            student_loss = ret_graph.var(ret_graph.out_nodes['loss'])._var
+            loss = distillation_loss + student_loss
+
+            ret_graph.out_nodes[
+                'soft_label_loss_' + self.student_feature_map + "_" +
+                self.teacher_feature_map] = distillation_loss.name
+            ret_graph.out_nodes['loss'] = loss.name
+        return ret_graph

python/paddle/fluid/contrib/slim/tests/distillation/compress.yaml

@@ -33,10 +33,17 @@ distillers:
         teacher_feature_map: 'teacher.tmp_2'
         student_feature_map: 'student.tmp_2'
         distillation_loss_weight: 1
+    soft_label_distiller:
+        class: 'SoftLabelDistiller'
+        student_temperature: 1.0
+        teacher_temperature: 1.0 
+        teacher_feature_map: 'teacher.tmp_1'
+        student_feature_map: 'student.tmp_1'
+        distillation_loss_weight: 0.001
 strategies:
     distillation_strategy:
         class: 'DistillationStrategy'
-        distillers: ['fsp_distiller', 'l2_distiller']
+        distillers: ['fsp_distiller', 'l2_distiller', 'soft_label_distiller']
         start_epoch: 0
         end_epoch: 1
 compressor: