Merge pull request #1042 from jo8hua/initializer_update

Updated deprecated tf.initialize_all_variables() with tf.global_varia…

autoencoder/autoencoder_models/Autoencoder.py

@@ -21,7 +21,7 @@ class Autoencoder(object):
         self.cost = 0.5 * tf.reduce_sum(tf.pow(tf.sub(self.reconstruction, self.x), 2.0))
         self.optimizer = optimizer.minimize(self.cost)
 
-        init = tf.initialize_all_variables()
+        init = tf.global_variables_initializer()
         self.sess = tf.Session()
         self.sess.run(init)
 

autoencoder/autoencoder_models/DenoisingAutoencoder.py

@@ -25,7 +25,7 @@ class AdditiveGaussianNoiseAutoencoder(object):
         self.cost = 0.5 * tf.reduce_sum(tf.pow(tf.sub(self.reconstruction, self.x), 2.0))
         self.optimizer = optimizer.minimize(self.cost)
 
-        init = tf.initialize_all_variables()
+        init = tf.global_variables_initializer()
         self.sess = tf.Session()
         self.sess.run(init)
 
@@ -92,7 +92,7 @@ class MaskingNoiseAutoencoder(object):
         self.cost = 0.5 * tf.reduce_sum(tf.pow(tf.sub(self.reconstruction, self.x), 2.0))
         self.optimizer = optimizer.minimize(self.cost)
 
-        init = tf.initialize_all_variables()
+        init = tf.global_variables_initializer()
         self.sess = tf.Session()
         self.sess.run(init)
 

autoencoder/autoencoder_models/VariationalAutoencoder.py

@@ -30,7 +30,7 @@ class VariationalAutoencoder(object):
         self.cost = tf.reduce_mean(reconstr_loss + latent_loss)
         self.optimizer = optimizer.minimize(self.cost)
 
-        init = tf.initialize_all_variables()
+        init = tf.global_variables_initializer()
         self.sess = tf.Session()
         self.sess.run(init)
 

differential_privacy/dp_sgd/dp_mnist/dp_mnist.py

@@ -344,7 +344,7 @@ def Train(mnist_train_file, mnist_test_file, network_parameters, num_steps,
     # We need to maintain the intialization sequence.
     for v in tf.trainable_variables():
       sess.run(tf.initialize_variables([v]))
-    sess.run(tf.initialize_all_variables())
+    sess.run(tf.global_variables_initializer())
     sess.run(init_ops)
 
     results = []

differential_privacy/multiple_teachers/deep_cnn.py

@@ -490,7 +490,7 @@ def train(images, labels, ckpt_path, dropout=False):
     print("Graph constructed and saver created")
 
     # Build an initialization operation to run below.
-    init = tf.initialize_all_variables()
+    init = tf.global_variables_initializer()
 
     # Create and init sessions
     sess = tf.Session(config=tf.ConfigProto(log_device_placement=FLAGS.log_device_placement)) #NOLINT(long-line)

differential_privacy/privacy_accountant/tf/accountant.py

@@ -124,7 +124,7 @@ class MomentsAccountant(object):
   """Privacy accountant which keeps track of moments of privacy loss.
 
   Note: The constructor of this class creates tf.Variables that must
-  be initialized with tf.initialize_all_variables() or similar calls.
+  be initialized with tf.global_variables_initializer() or similar calls.
 
   MomentsAccountant accumulates the high moments of the privacy loss. It
   requires a method for computing differenital moments of the noise (See

inception/inception/inception_distributed_train.py

@@ -236,7 +236,7 @@ def train(target, dataset, cluster_spec):
       summary_op = tf.merge_all_summaries()
 
       # Build an initialization operation to run below.
-      init_op = tf.initialize_all_variables()
+      init_op = tf.global_variables_initializer()
 
       # We run the summaries in the same thread as the training operations by
       # passing in None for summary_op to avoid a summary_thread being started.

inception/inception/inception_train.py

@@ -307,7 +307,7 @@ def train(dataset):
     summary_op = tf.merge_summary(summaries)
 
     # Build an initialization operation to run below.
-    init = tf.initialize_all_variables()
+    init = tf.global_variables_initializer()
 
     # Start running operations on the Graph. allow_soft_placement must be set to
     # True to build towers on GPU, as some of the ops do not have GPU

inception/inception/slim/inception_test.py

@@ -95,7 +95,7 @@ class InceptionTest(tf.test.TestCase):
       self.assertListEqual(logits.get_shape().as_list(),
                            [None, num_classes])
       images = tf.random_uniform((batch_size, height, width, 3))
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits, {inputs: images.eval()})
       self.assertEquals(output.shape, (batch_size, num_classes))
 
@@ -108,7 +108,7 @@ class InceptionTest(tf.test.TestCase):
       logits, _ = inception.inception_v3(eval_inputs, num_classes,
                                          is_training=False)
       predictions = tf.argmax(logits, 1)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (batch_size,))
 
@@ -125,7 +125,7 @@ class InceptionTest(tf.test.TestCase):
       logits, _ = inception.inception_v3(eval_inputs, num_classes,
                                          is_training=False)
       predictions = tf.argmax(logits, 1)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (eval_batch_size,))
 

inception/inception/slim/ops_test.py

@@ -128,7 +128,7 @@ class ConvTest(tf.test.TestCase):
       wd = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)[0]
       self.assertEquals(wd.op.name,
                         'Conv/weights/Regularizer/L2Regularizer/value')
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       self.assertTrue(sess.run(wd) <= 0.01)
 
   def testCreateConvWithoutWD(self):
@@ -254,7 +254,7 @@ class FCTest(tf.test.TestCase):
       wd = tf.get_collection(tf.GraphKeys.REGULARIZATION_LOSSES)[0]
       self.assertEquals(wd.op.name,
                         'FC/weights/Regularizer/L2Regularizer/value')
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       self.assertTrue(sess.run(wd) <= 0.01)
 
   def testCreateFCWithoutWD(self):
@@ -604,7 +604,7 @@ class BatchNormTest(tf.test.TestCase):
         barrier = tf.no_op(name='gradient_barrier')
         output = control_flow_ops.with_dependencies([barrier], output)
       # Initialize all variables
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       moving_mean = variables.get_variables('BatchNorm/moving_mean')[0]
       moving_variance = variables.get_variables('BatchNorm/moving_variance')[0]
       mean, variance = sess.run([moving_mean, moving_variance])
@@ -634,7 +634,7 @@ class BatchNormTest(tf.test.TestCase):
         barrier = tf.no_op(name='gradient_barrier')
         output = control_flow_ops.with_dependencies([barrier], output)
       # Initialize all variables
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       moving_mean = variables.get_variables('BatchNorm/moving_mean')[0]
       moving_variance = variables.get_variables('BatchNorm/moving_variance')[0]
       mean, variance = sess.run([moving_mean, moving_variance])
@@ -668,7 +668,7 @@ class BatchNormTest(tf.test.TestCase):
         barrier = tf.no_op(name='gradient_barrier')
         output = control_flow_ops.with_dependencies([barrier], output)
       # Initialize all variables
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       moving_mean = variables.get_variables('BatchNorm/moving_mean')[0]
       moving_variance = variables.get_variables('BatchNorm/moving_variance')[0]
       mean, variance = sess.run([moving_mean, moving_variance])

namignizer/names.py

@@ -160,7 +160,7 @@ def train(data_dir, checkpoint_path, config):
         with tf.variable_scope("model", reuse=None, initializer=initializer):
             m = NamignizerModel(is_training=True, config=config)
 
-        tf.initialize_all_variables().run()
+        tf.global_variables_initializer().run()
 
         for i in range(config.max_max_epoch):
             lr_decay = config.lr_decay ** max(i - config.max_epoch, 0.0)

neural_gpu/neural_gpu_trainer.py

@@ -377,7 +377,7 @@ def initialize(sess=None):
                    % ckpt.model_checkpoint_path)
     model.saver.restore(sess, ckpt.model_checkpoint_path)
   elif sv is None:
-    sess.run(tf.initialize_all_variables())
+    sess.run(tf.global_variables_initializer())
     data.print_out("Initialized variables (no supervisor mode).")
   elif FLAGS.task < 1 and FLAGS.mem_size > 0:
     # sess.run(model.mem_norm_op)

neural_programmer/model.py

@@ -674,5 +674,5 @@ class Graph():
         use_locking=True)
     self.step = adam.apply_gradients(zip(grads, optimize_params), 
 					global_step=self.global_step)
-    self.init_op = tf.initialize_all_variables()
+    self.init_op = tf.global_variables_initializer()
 

slim/deployment/model_deploy_test.py

@@ -502,7 +502,7 @@ class DeployTest(tf.test.TestCase):
       self.assertEqual(model.train_op.op.name, 'train_op')
 
       with tf.Session() as sess:
-        sess.run(tf.initialize_all_variables())
+        sess.run(tf.global_variables_initializer())
         moving_mean = tf.contrib.framework.get_variables_by_name(
             'moving_mean')[0]
         moving_variance = tf.contrib.framework.get_variables_by_name(

slim/nets/alexnet_test.py

@@ -137,7 +137,7 @@ class AlexnetV2Test(tf.test.TestCase):
     with self.test_session() as sess:
       inputs = tf.random_uniform((batch_size, height, width, 3))
       logits, _ = alexnet.alexnet_v2(inputs)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits)
       self.assertTrue(output.any())
 

slim/nets/inception_resnet_v2_test.py

@@ -95,7 +95,7 @@ class InceptionTest(tf.test.TestCase):
       self.assertListEqual(logits.get_shape().as_list(),
                            [None, num_classes])
       images = tf.random_uniform((batch_size, height, width, 3))
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits, {inputs: images.eval()})
       self.assertEquals(output.shape, (batch_size, num_classes))
 
@@ -109,7 +109,7 @@ class InceptionTest(tf.test.TestCase):
                                                 num_classes,
                                                 is_training=False)
       predictions = tf.argmax(logits, 1)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (batch_size,))
 
@@ -127,7 +127,7 @@ class InceptionTest(tf.test.TestCase):
                                                 is_training=False,
                                                 reuse=True)
       predictions = tf.argmax(logits, 1)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (eval_batch_size,))
 

slim/nets/inception_v1_test.py

@@ -140,7 +140,7 @@ class InceptionV1Test(tf.test.TestCase):
                            [batch_size, num_classes])
       pre_pool = end_points['Mixed_5c']
       feed_dict = {inputs: input_np}
-      tf.initialize_all_variables().run()
+      tf.global_variables_initializer().run()
       pre_pool_out = sess.run(pre_pool, feed_dict=feed_dict)
       self.assertListEqual(list(pre_pool_out.shape), [batch_size, 7, 7, 1024])
 
@@ -157,7 +157,7 @@ class InceptionV1Test(tf.test.TestCase):
     images = tf.random_uniform((batch_size, height, width, 3))
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits, {inputs: images.eval()})
       self.assertEquals(output.shape, (batch_size, num_classes))
 
@@ -172,7 +172,7 @@ class InceptionV1Test(tf.test.TestCase):
     predictions = tf.argmax(logits, 1)
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (batch_size,))
 
@@ -189,7 +189,7 @@ class InceptionV1Test(tf.test.TestCase):
     predictions = tf.argmax(logits, 1)
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (eval_batch_size,))
 
@@ -201,7 +201,7 @@ class InceptionV1Test(tf.test.TestCase):
                                        spatial_squeeze=False)
 
     with self.test_session() as sess:
-      tf.initialize_all_variables().run()
+      tf.global_variables_initializer().run()
       logits_out = sess.run(logits)
       self.assertListEqual(list(logits_out.shape), [1, 1, 1, num_classes])
 

slim/nets/inception_v2_test.py

@@ -192,7 +192,7 @@ class InceptionV2Test(tf.test.TestCase):
                            [batch_size, num_classes])
       pre_pool = end_points['Mixed_5c']
       feed_dict = {inputs: input_np}
-      tf.initialize_all_variables().run()
+      tf.global_variables_initializer().run()
       pre_pool_out = sess.run(pre_pool, feed_dict=feed_dict)
       self.assertListEqual(list(pre_pool_out.shape), [batch_size, 7, 7, 1024])
 
@@ -209,7 +209,7 @@ class InceptionV2Test(tf.test.TestCase):
     images = tf.random_uniform((batch_size, height, width, 3))
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits, {inputs: images.eval()})
       self.assertEquals(output.shape, (batch_size, num_classes))
 
@@ -224,7 +224,7 @@ class InceptionV2Test(tf.test.TestCase):
     predictions = tf.argmax(logits, 1)
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (batch_size,))
 
@@ -241,7 +241,7 @@ class InceptionV2Test(tf.test.TestCase):
     predictions = tf.argmax(logits, 1)
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (eval_batch_size,))
 
@@ -253,7 +253,7 @@ class InceptionV2Test(tf.test.TestCase):
                                        spatial_squeeze=False)
 
     with self.test_session() as sess:
-      tf.initialize_all_variables().run()
+      tf.global_variables_initializer().run()
       logits_out = sess.run(logits)
       self.assertListEqual(list(logits_out.shape), [1, 1, 1, num_classes])
 

slim/nets/inception_v3_test.py

@@ -221,7 +221,7 @@ class InceptionV3Test(tf.test.TestCase):
                            [batch_size, num_classes])
       pre_pool = end_points['Mixed_7c']
       feed_dict = {inputs: input_np}
-      tf.initialize_all_variables().run()
+      tf.global_variables_initializer().run()
       pre_pool_out = sess.run(pre_pool, feed_dict=feed_dict)
       self.assertListEqual(list(pre_pool_out.shape), [batch_size, 8, 8, 2048])
 
@@ -238,7 +238,7 @@ class InceptionV3Test(tf.test.TestCase):
     images = tf.random_uniform((batch_size, height, width, 3))
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits, {inputs: images.eval()})
       self.assertEquals(output.shape, (batch_size, num_classes))
 
@@ -253,7 +253,7 @@ class InceptionV3Test(tf.test.TestCase):
     predictions = tf.argmax(logits, 1)
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (batch_size,))
 
@@ -271,7 +271,7 @@ class InceptionV3Test(tf.test.TestCase):
     predictions = tf.argmax(logits, 1)
 
     with self.test_session() as sess:
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (eval_batch_size,))
 
@@ -283,7 +283,7 @@ class InceptionV3Test(tf.test.TestCase):
                                        spatial_squeeze=False)
 
     with self.test_session() as sess:
-      tf.initialize_all_variables().run()
+      tf.global_variables_initializer().run()
       logits_out = sess.run(logits)
       self.assertListEqual(list(logits_out.shape), [1, 1, 1, num_classes])
 

slim/nets/inception_v4_test.py

@@ -175,7 +175,7 @@ class InceptionTest(tf.test.TestCase):
       self.assertListEqual(logits.get_shape().as_list(),
                            [None, num_classes])
       images = tf.random_uniform((batch_size, height, width, 3))
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(logits, {inputs: images.eval()})
       self.assertEquals(output.shape, (batch_size, num_classes))
 
@@ -189,7 +189,7 @@ class InceptionTest(tf.test.TestCase):
                                          num_classes,
                                          is_training=False)
       predictions = tf.argmax(logits, 1)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (batch_size,))
 
@@ -207,7 +207,7 @@ class InceptionTest(tf.test.TestCase):
                                          is_training=False,
                                          reuse=True)
       predictions = tf.argmax(logits, 1)
-      sess.run(tf.initialize_all_variables())
+      sess.run(tf.global_variables_initializer())
       output = sess.run(predictions)
       self.assertEquals(output.shape, (eval_batch_size,))