Question

我正在尝试在tensor-flow中实现批量规范化层。使用tf.moments来运行此列车的步骤没有问题，以获得平均值和方差。

对于测试时间，我想设置一个指数移动平均线来跟踪均值和方差。我想这样做：

def batch_normalized_linear_layer(state_below, scope_name, n_inputs, n_outputs, stddev, wd, eps=.0001):
  with tf.variable_scope(scope_name) as scope:
    weight = _variable_with_weight_decay(
      "weights", shape=[n_inputs, n_outputs],
      stddev=stddev, wd=wd
    )
    act = tf.matmul(state_below, weight)
    # get moments
    act_mean, act_variance = tf.nn.moments(act, [0])
    # get mean and variance variables
    mean = _variable_on_cpu('bn_mean', [n_outputs], tf.constant_initializer(0.0))
    variance = _variable_on_cpu('bn_variance', [n_outputs], tf.constant_initializer(1.0))
    # assign the moments
    assign_mean = mean.assign(act_mean)
    assign_variance = variance.assign(act_variance)

    act_bn = tf.mul((act - mean), tf.rsqrt(variance + eps), name=scope.name+"_bn")

    beta = _variable_on_cpu("beta", [n_outputs], tf.constant_initializer(0.0))
    gamma = _variable_on_cpu("gamma", [n_outputs], tf.constant_initializer(1.0))
    bn = tf.add(tf.mul(act_bn, gamma), beta)
    output = tf.nn.relu(bn, name=scope.name)
    _activation_summary(output)
  return output, mean, variance

_variable_on_cpu定义为：

def _variable_on_cpu(name, shape, initializer):
  """Helper to create a Variable stored on CPU memory.

  Args:
    name: name of the variable
    shape: list of ints
    initializer: initializer for Variable

  Returns:
    Variable Tensor
  """
  with tf.device('/cpu:0'):
    var = tf.get_variable(name, shape, initializer=initializer)
  return var

我相信我正在设置

assign_mean = mean.assign(act_mean)
assign_variance = variance.assign(act_variance)

错误，但我不确定如何。当我使用张量板来跟踪这些均值和方差变量时，它们只是与它们的初始值相等。

Answer 1

Rafal的评论是问题的核心：您没有运行分配节点。您可以尝试使用我在另一个答案中发布的batchnorm助手 - How could I use Batch Normalization in TensorFlow? - 或者您可以通过添加with_dependencies强制分配，正如他所建议的那样。

一般原则是，如果数据或控制依赖关系流过＆＃34;通过＆＃34;您应该只依赖于正在运行的节点。它。 with_dependencies确保在使用输出op之前，指定的依赖项已经完成。

使用张量流实现批量标准化

1 个答案: