small fix on keras examples

examples/DeepQNetwork/DQNModel.py

@@ -83,8 +83,7 @@ class Model(ModelDesc):
     def optimizer(self):
         lr = tf.get_variable('learning_rate', initializer=self.learning_rate, trainable=False)
         opt = tf.train.AdamOptimizer(lr, epsilon=1e-3)
-        return optimizer.apply_grad_processors(
-            opt, [gradproc.GlobalNormClip(10), gradproc.SummaryGradient()])
+        return optimizer.apply_grad_processors(opt, [gradproc.SummaryGradient()])
 
     @staticmethod
     def update_target_param():

examples/keras/imagenet-resnet-keras.py

@@ -4,6 +4,7 @@
 # Author: Yuxin Wu
 
 import numpy as np
+import os
 import tensorflow as tf
 import argparse
 

tensorpack/contrib/keras.py

@@ -56,6 +56,7 @@ class KerasModelCaller(object):
 
         old_trainable_names = set([x.name for x in tf.trainable_variables()])
         trainable_backup = backup_collection([tf.GraphKeys.TRAINABLE_VARIABLES])
+        update_ops_backup = backup_collection([tf.GraphKeys.UPDATE_OPS])
 
         def post_process_model(model):
             added_trainable_names = set([x.name for x in tf.trainable_variables()])
@@ -73,6 +74,11 @@ class KerasModelCaller(object):
                     logger.warn("Keras created trainable variable '{}' which is actually not trainable. "
                                 "This was automatically corrected by tensorpack.".format(n))
 
+            # Keras models might not use this collection at all (in some versions).
+            restore_collection(update_ops_backup)
+            for op in model.updates:
+                tf.add_to_collection(tf.GraphKeys.UPDATE_OPS, op)
+
         if self.cached_model is None:
             assert not reuse
             model = self.cached_model = self.get_model(*input_tensors)