MCPcopy Create free account
hub / github.com/DeepRec-AI/DeepRec / batch_norm_with_global_normalization

Function batch_norm_with_global_normalization

tensorflow/python/ops/nn_impl.py:1655–1698  ·  view source on GitHub ↗

Batch normalization. This op is deprecated. See `tf.nn.batch_normalization`. Args: t: A 4D input Tensor. m: A 1D mean Tensor with size matching the last dimension of t. This is the first output from tf.nn.moments, or a saved moving average thereof. v: A 1D variance Tens

(t=None,
                                         m=None,
                                         v=None,
                                         beta=None,
                                         gamma=None,
                                         variance_epsilon=None,
                                         scale_after_normalization=None,
                                         name=None,
                                         input=None,  # pylint: disable=redefined-builtin
                                         mean=None,
                                         variance=None)

Source from the content-addressed store, hash-verified

1653
1654@tf_export(v1=["nn.batch_norm_with_global_normalization"])
1655def batch_norm_with_global_normalization(t=None,
1656 m=None,
1657 v=None,
1658 beta=None,
1659 gamma=None,
1660 variance_epsilon=None,
1661 scale_after_normalization=None,
1662 name=None,
1663 input=None, # pylint: disable=redefined-builtin
1664 mean=None,
1665 variance=None):
1666 """Batch normalization.
1667
1668 This op is deprecated. See `tf.nn.batch_normalization`.
1669
1670 Args:
1671 t: A 4D input Tensor.
1672 m: A 1D mean Tensor with size matching the last dimension of t.
1673 This is the first output from tf.nn.moments,
1674 or a saved moving average thereof.
1675 v: A 1D variance Tensor with size matching the last dimension of t.
1676 This is the second output from tf.nn.moments,
1677 or a saved moving average thereof.
1678 beta: A 1D beta Tensor with size matching the last dimension of t.
1679 An offset to be added to the normalized tensor.
1680 gamma: A 1D gamma Tensor with size matching the last dimension of t.
1681 If "scale_after_normalization" is true, this tensor will be multiplied
1682 with the normalized tensor.
1683 variance_epsilon: A small float number to avoid dividing by 0.
1684 scale_after_normalization: A bool indicating whether the resulted tensor
1685 needs to be multiplied with gamma.
1686 name: A name for this operation (optional).
1687 input: Alias for t.
1688 mean: Alias for m.
1689 variance: Alias for v.
1690
1691 Returns:
1692 A batch-normalized `t`.
1693 """
1694 t = deprecated_argument_lookup("input", input, "t", t)
1695 m = deprecated_argument_lookup("mean", mean, "m", m)
1696 v = deprecated_argument_lookup("variance", variance, "v", v)
1697 return batch_normalization(t, m, v, beta, gamma if scale_after_normalization
1698 else None, variance_epsilon, name)
1699
1700
1701# pylint: disable=redefined-builtin,line-too-long

Calls 2

batch_normalizationFunction · 0.70

Tested by

no test coverage detected