Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

How could I use batch normalization in TensorFlow?

I would like to use batch normalization in TensorFlow. I found the related C++ source code in core/ops/nn_ops.cc. However, I did not find it documented on tensorflow.org.

BN has different semantics in MLP and CNN, so I am not sure what exactly this BN does.

I did not find a method called MovingMoments either.

like image 958
Shawn Lee Avatar asked Nov 27 '15 03:11

Shawn Lee


People also ask

What is batch normalization in Tensorflow?

Batch normalization applies a transformation that maintains the mean output close to 0 and the output standard deviation close to 1. Importantly, batch normalization works differently during training and during inference.

How do I batch normalize in keras?

Batch Normalization in the Keras API Put simply, Batch Normalization can be added as easily as adding a BatchNormalization() layer to your model, e.g. with model. add . However, if you wish, local parameters can be tuned to steer the way in which Batch Normalization works.

How do I batch normalize in CNN?

Batch Norm is a normalization technique done between the layers of a Neural Network instead of in the raw data. It is done along mini-batches instead of the full data set. It serves to speed up training and use higher learning rates, making learning easier. the standard deviation of the neurons' output.

Where do we use normalization in batch?

When to use Batch Normalization? We can use Batch Normalization in Convolution Neural Networks, Recurrent Neural Networks, and Artificial Neural Networks. In practical coding, we add Batch Normalization after the activation function of the output layer or before the activation function of the input layer.


7 Answers

Update July 2016 The easiest way to use batch normalization in TensorFlow is through the higher-level interfaces provided in either contrib/layers, tflearn, or slim.

Previous answer if you want to DIY: The documentation string for this has improved since the release - see the docs comment in the master branch instead of the one you found. It clarifies, in particular, that it's the output from tf.nn.moments.

You can see a very simple example of its use in the batch_norm test code. For a more real-world use example, I've included below the helper class and use notes that I scribbled up for my own use (no warranty provided!):

"""A helper class for managing batch normalization state.                   

This class is designed to simplify adding batch normalization               
(http://arxiv.org/pdf/1502.03167v3.pdf) to your model by                    
managing the state variables associated with it.                            

Important use note:  The function get_assigner() returns                    
an op that must be executed to save the updated state.                      
A suggested way to do this is to make execution of the                      
model optimizer force it, e.g., by:                                         

  update_assignments = tf.group(bn1.get_assigner(),                         
                                bn2.get_assigner())                         
  with tf.control_dependencies([optimizer]):                                
    optimizer = tf.group(update_assignments)                                

"""

import tensorflow as tf


class ConvolutionalBatchNormalizer(object):
  """Helper class that groups the normalization logic and variables.        

  Use:                                                                      
      ewma = tf.train.ExponentialMovingAverage(decay=0.99)                  
      bn = ConvolutionalBatchNormalizer(depth, 0.001, ewma, True)           
      update_assignments = bn.get_assigner()                                
      x = bn.normalize(y, train=training?)                                  
      (the output x will be batch-normalized).                              
  """

  def __init__(self, depth, epsilon, ewma_trainer, scale_after_norm):
    self.mean = tf.Variable(tf.constant(0.0, shape=[depth]),
                            trainable=False)
    self.variance = tf.Variable(tf.constant(1.0, shape=[depth]),
                                trainable=False)
    self.beta = tf.Variable(tf.constant(0.0, shape=[depth]))
    self.gamma = tf.Variable(tf.constant(1.0, shape=[depth]))
    self.ewma_trainer = ewma_trainer
    self.epsilon = epsilon
    self.scale_after_norm = scale_after_norm

  def get_assigner(self):
    """Returns an EWMA apply op that must be invoked after optimization."""
    return self.ewma_trainer.apply([self.mean, self.variance])

  def normalize(self, x, train=True):
    """Returns a batch-normalized version of x."""
    if train:
      mean, variance = tf.nn.moments(x, [0, 1, 2])
      assign_mean = self.mean.assign(mean)
      assign_variance = self.variance.assign(variance)
      with tf.control_dependencies([assign_mean, assign_variance]):
        return tf.nn.batch_norm_with_global_normalization(
            x, mean, variance, self.beta, self.gamma,
            self.epsilon, self.scale_after_norm)
    else:
      mean = self.ewma_trainer.average(self.mean)
      variance = self.ewma_trainer.average(self.variance)
      local_beta = tf.identity(self.beta)
      local_gamma = tf.identity(self.gamma)
      return tf.nn.batch_norm_with_global_normalization(
          x, mean, variance, local_beta, local_gamma,
          self.epsilon, self.scale_after_norm)

Note that I called it a ConvolutionalBatchNormalizer because it pins the use of tf.nn.moments to sum across axes 0, 1, and 2, whereas for non-convolutional use you might only want axis 0.

Feedback appreciated if you use it.

like image 161
dga Avatar answered Oct 13 '22 04:10

dga


As of TensorFlow 1.0 (February 2017) there's also the high-level tf.layers.batch_normalization API included in TensorFlow itself.

It's super simple to use:

# Set this to True for training and False for testing
training = tf.placeholder(tf.bool)

x = tf.layers.dense(input_x, units=100)
x = tf.layers.batch_normalization(x, training=training)
x = tf.nn.relu(x)

...except that it adds extra ops to the graph (for updating its mean and variance variables) in such a way that they won't be dependencies of your training op. You can either just run the ops separately:

extra_update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
sess.run([train_op, extra_update_ops], ...)

or add the update ops as dependencies of your training op manually, then just run your training op as normal:

extra_update_ops = tf.get_collection(tf.GraphKeys.UPDATE_OPS)
with tf.control_dependencies(extra_update_ops):
    train_op = optimizer.minimize(loss)
...
sess.run([train_op], ...)
like image 20
Matthew Rahtz Avatar answered Oct 13 '22 05:10

Matthew Rahtz


The following works fine for me, it does not require invoking EMA-apply outside.

import numpy as np
import tensorflow as tf
from tensorflow.python import control_flow_ops

def batch_norm(x, n_out, phase_train, scope='bn'):
    """
    Batch normalization on convolutional maps.
    Args:
        x:           Tensor, 4D BHWD input maps
        n_out:       integer, depth of input maps
        phase_train: boolean tf.Varialbe, true indicates training phase
        scope:       string, variable scope
    Return:
        normed:      batch-normalized maps
    """
    with tf.variable_scope(scope):
        beta = tf.Variable(tf.constant(0.0, shape=[n_out]),
                                     name='beta', trainable=True)
        gamma = tf.Variable(tf.constant(1.0, shape=[n_out]),
                                      name='gamma', trainable=True)
        batch_mean, batch_var = tf.nn.moments(x, [0,1,2], name='moments')
        ema = tf.train.ExponentialMovingAverage(decay=0.5)

        def mean_var_with_update():
            ema_apply_op = ema.apply([batch_mean, batch_var])
            with tf.control_dependencies([ema_apply_op]):
                return tf.identity(batch_mean), tf.identity(batch_var)

        mean, var = tf.cond(phase_train,
                            mean_var_with_update,
                            lambda: (ema.average(batch_mean), ema.average(batch_var)))
        normed = tf.nn.batch_normalization(x, mean, var, beta, gamma, 1e-3)
    return normed

Example:

import math

n_in, n_out = 3, 16
ksize = 3
stride = 1
phase_train = tf.placeholder(tf.bool, name='phase_train')
input_image = tf.placeholder(tf.float32, name='input_image')
kernel = tf.Variable(tf.truncated_normal([ksize, ksize, n_in, n_out],
                                   stddev=math.sqrt(2.0/(ksize*ksize*n_out))),
                                   name='kernel')
conv = tf.nn.conv2d(input_image, kernel, [1,stride,stride,1], padding='SAME')
conv_bn = batch_norm(conv, n_out, phase_train)
relu = tf.nn.relu(conv_bn)

with tf.Session() as session:
    session.run(tf.initialize_all_variables())
    for i in range(20):
        test_image = np.random.rand(4,32,32,3)
        sess_outputs = session.run([relu],
          {input_image.name: test_image, phase_train.name: True})
like image 22
bgshi Avatar answered Oct 13 '22 06:10

bgshi


There is also an "official" batch normalization layer coded by the developers. They don't have very good docs on how to use it but here is how to use it (according to me):

from tensorflow.contrib.layers.python.layers import batch_norm as batch_norm

def batch_norm_layer(x,train_phase,scope_bn):
    bn_train = batch_norm(x, decay=0.999, center=True, scale=True,
    updates_collections=None,
    is_training=True,
    reuse=None, # is this right?
    trainable=True,
    scope=scope_bn)
    bn_inference = batch_norm(x, decay=0.999, center=True, scale=True,
    updates_collections=None,
    is_training=False,
    reuse=True, # is this right?
    trainable=True,
    scope=scope_bn)
    z = tf.cond(train_phase, lambda: bn_train, lambda: bn_inference)
    return z

to actually use it you need to create a placeholder for train_phase that indicates if you are in training or inference phase (as in train_phase = tf.placeholder(tf.bool, name='phase_train')). Its value can be filled during inference or training with a tf.session as in:

test_error = sess.run(fetches=cross_entropy, feed_dict={x: batch_xtest, y_:batch_ytest, train_phase: False})

or during training:

sess.run(fetches=train_step, feed_dict={x: batch_xs, y_:batch_ys, train_phase: True})

I'm pretty sure this is correct according to the discussion in github.


Seems there is another useful link:

http://r2rt.com/implementing-batch-normalization-in-tensorflow.html

like image 32
Charlie Parker Avatar answered Oct 13 '22 06:10

Charlie Parker


You can simply use the build-in batch_norm layer:

batch_norm = tf.cond(is_train, 
    lambda: tf.contrib.layers.batch_norm(prev, activation_fn=tf.nn.relu, is_training=True, reuse=None),
    lambda: tf.contrib.layers.batch_norm(prev, activation_fn =tf.nn.relu, is_training=False, reuse=True))

where prev is the output of your previous layer (can be both fully-connected or a convolutional layer) and is_train is a boolean placeholder. Just use batch_norm as the input to the next layer, then.

like image 33
Martina Marek Avatar answered Oct 13 '22 05:10

Martina Marek


Since someone recently edited this, I'd like to clarify that this is no longer an issue.

This answer does not seem correct When phase_train is set to false, it still updates the ema mean and variance. This can be verified with the following code snippet.

x = tf.placeholder(tf.float32, [None, 20, 20, 10], name='input')
phase_train = tf.placeholder(tf.bool, name='phase_train')

# generate random noise to pass into batch norm
x_gen = tf.random_normal([50,20,20,10])
pt_false = tf.Variable(tf.constant(True))

#generate a constant variable to pass into batch norm
y = x_gen.eval()

[bn, bn_vars] = batch_norm(x, 10, phase_train)

tf.initialize_all_variables().run()
train_step = lambda: bn.eval({x:x_gen.eval(), phase_train:True})
test_step = lambda: bn.eval({x:y, phase_train:False})
test_step_c = lambda: bn.eval({x:y, phase_train:True})

# Verify that this is different as expected, two different x's have different norms
print(train_step()[0][0][0])
print(train_step()[0][0][0])

# Verify that this is same as expected, same x's (y) have same norm
print(train_step_c()[0][0][0])
print(train_step_c()[0][0][0])

# THIS IS DIFFERENT but should be they same, should only be reading from the ema.
print(test_step()[0][0][0])
print(test_step()[0][0][0])
like image 31
jrock Avatar answered Oct 13 '22 04:10

jrock


Using TensorFlow built-in batch_norm layer, below is the code to load data, build a network with one hidden ReLU layer and L2 normalization and introduce batch normalization for both hidden and out layer. This runs fine and trains fine. Just FYI this example is mostly built upon the data and code from Udacity DeepLearning course. P.S. Yes, parts of it were discussed one way or another in answers earlier but I decided to gather in one code snippet everything so that you have example of whole network training process with Batch Normalization and its evaluation

# These are all the modules we'll be using later. Make sure you can import them
# before proceeding further.
from __future__ import print_function
import numpy as np
import tensorflow as tf
from six.moves import cPickle as pickle

pickle_file = '/home/maxkhk/Documents/Udacity/DeepLearningCourse/SourceCode/tensorflow/examples/udacity/notMNIST.pickle'

with open(pickle_file, 'rb') as f:
  save = pickle.load(f)
  train_dataset = save['train_dataset']
  train_labels = save['train_labels']
  valid_dataset = save['valid_dataset']
  valid_labels = save['valid_labels']
  test_dataset = save['test_dataset']
  test_labels = save['test_labels']
  del save  # hint to help gc free up memory
  print('Training set', train_dataset.shape, train_labels.shape)
  print('Validation set', valid_dataset.shape, valid_labels.shape)
  print('Test set', test_dataset.shape, test_labels.shape)

image_size = 28
num_labels = 10

def reformat(dataset, labels):
  dataset = dataset.reshape((-1, image_size * image_size)).astype(np.float32)
  # Map 2 to [0.0, 1.0, 0.0 ...], 3 to [0.0, 0.0, 1.0 ...]
  labels = (np.arange(num_labels) == labels[:,None]).astype(np.float32)
  return dataset, labels
train_dataset, train_labels = reformat(train_dataset, train_labels)
valid_dataset, valid_labels = reformat(valid_dataset, valid_labels)
test_dataset, test_labels = reformat(test_dataset, test_labels)
print('Training set', train_dataset.shape, train_labels.shape)
print('Validation set', valid_dataset.shape, valid_labels.shape)
print('Test set', test_dataset.shape, test_labels.shape)


def accuracy(predictions, labels):
  return (100.0 * np.sum(np.argmax(predictions, 1) == np.argmax(labels, 1))
          / predictions.shape[0])


#for NeuralNetwork model code is below
#We will use SGD for training to save our time. Code is from Assignment 2
#beta is the new parameter - controls level of regularization.
#Feel free to play with it - the best one I found is 0.001
#notice, we introduce L2 for both biases and weights of all layers

batch_size = 128
beta = 0.001

#building tensorflow graph
graph = tf.Graph()
with graph.as_default():
      # Input data. For the training data, we use a placeholder that will be fed
  # at run time with a training minibatch.
  tf_train_dataset = tf.placeholder(tf.float32,
                                    shape=(batch_size, image_size * image_size))
  tf_train_labels = tf.placeholder(tf.float32, shape=(batch_size, num_labels))
  tf_valid_dataset = tf.constant(valid_dataset)
  tf_test_dataset = tf.constant(test_dataset)

  #introduce batchnorm
  tf_train_dataset_bn = tf.contrib.layers.batch_norm(tf_train_dataset)


  #now let's build our new hidden layer
  #that's how many hidden neurons we want
  num_hidden_neurons = 1024
  #its weights
  hidden_weights = tf.Variable(
    tf.truncated_normal([image_size * image_size, num_hidden_neurons]))
  hidden_biases = tf.Variable(tf.zeros([num_hidden_neurons]))

  #now the layer itself. It multiplies data by weights, adds biases
  #and takes ReLU over result
  hidden_layer = tf.nn.relu(tf.matmul(tf_train_dataset_bn, hidden_weights) + hidden_biases)

  #adding the batch normalization layerhi()
  hidden_layer_bn = tf.contrib.layers.batch_norm(hidden_layer)

  #time to go for output linear layer
  #out weights connect hidden neurons to output labels
  #biases are added to output labels  
  out_weights = tf.Variable(
    tf.truncated_normal([num_hidden_neurons, num_labels]))  

  out_biases = tf.Variable(tf.zeros([num_labels]))  

  #compute output  
  out_layer = tf.matmul(hidden_layer_bn,out_weights) + out_biases
  #our real output is a softmax of prior result
  #and we also compute its cross-entropy to get our loss
  #Notice - we introduce our L2 here
  loss = (tf.reduce_mean(tf.nn.softmax_cross_entropy_with_logits(
    out_layer, tf_train_labels) +
    beta*tf.nn.l2_loss(hidden_weights) +
    beta*tf.nn.l2_loss(hidden_biases) +
    beta*tf.nn.l2_loss(out_weights) +
    beta*tf.nn.l2_loss(out_biases)))

  #now we just minimize this loss to actually train the network
  optimizer = tf.train.GradientDescentOptimizer(0.5).minimize(loss)

  #nice, now let's calculate the predictions on each dataset for evaluating the
  #performance so far
  # Predictions for the training, validation, and test data.
  train_prediction = tf.nn.softmax(out_layer)
  valid_relu = tf.nn.relu(  tf.matmul(tf_valid_dataset, hidden_weights) + hidden_biases)
  valid_prediction = tf.nn.softmax( tf.matmul(valid_relu, out_weights) + out_biases) 

  test_relu = tf.nn.relu( tf.matmul( tf_test_dataset, hidden_weights) + hidden_biases)
  test_prediction = tf.nn.softmax(tf.matmul(test_relu, out_weights) + out_biases)



#now is the actual training on the ANN we built
#we will run it for some number of steps and evaluate the progress after 
#every 500 steps

#number of steps we will train our ANN
num_steps = 3001

#actual training
with tf.Session(graph=graph) as session:
  tf.initialize_all_variables().run()
  print("Initialized")
  for step in range(num_steps):
    # Pick an offset within the training data, which has been randomized.
    # Note: we could use better randomization across epochs.
    offset = (step * batch_size) % (train_labels.shape[0] - batch_size)
    # Generate a minibatch.
    batch_data = train_dataset[offset:(offset + batch_size), :]
    batch_labels = train_labels[offset:(offset + batch_size), :]
    # Prepare a dictionary telling the session where to feed the minibatch.
    # The key of the dictionary is the placeholder node of the graph to be fed,
    # and the value is the numpy array to feed to it.
    feed_dict = {tf_train_dataset : batch_data, tf_train_labels : batch_labels}
    _, l, predictions = session.run(
      [optimizer, loss, train_prediction], feed_dict=feed_dict)
    if (step % 500 == 0):
      print("Minibatch loss at step %d: %f" % (step, l))
      print("Minibatch accuracy: %.1f%%" % accuracy(predictions, batch_labels))
      print("Validation accuracy: %.1f%%" % accuracy(
        valid_prediction.eval(), valid_labels))
      print("Test accuracy: %.1f%%" % accuracy(test_prediction.eval(), test_labels))
like image 21
Maksim Khaitovich Avatar answered Oct 13 '22 05:10

Maksim Khaitovich