From a762b0a0de067036dc799ac37127142a9ac2fed3 Mon Sep 17 00:00:00 2001 From: "A. Unique TensorFlower" Date: Tue, 10 Jan 2017 12:59:31 -0800 Subject: [PATCH] Update generated Python Op docs. Change: 144117353 --- .../python/functions_and_classes/shard1/tf.nn.moments.md | 8 ++++++-- tensorflow/g3doc/api_docs/python/nn.md | 8 ++++++-- 2 files changed, 12 insertions(+), 4 deletions(-) diff --git a/tensorflow/g3doc/api_docs/python/functions_and_classes/shard1/tf.nn.moments.md b/tensorflow/g3doc/api_docs/python/functions_and_classes/shard1/tf.nn.moments.md index d3aa88d68de..dd56055311e 100644 --- a/tensorflow/g3doc/api_docs/python/functions_and_classes/shard1/tf.nn.moments.md +++ b/tensorflow/g3doc/api_docs/python/functions_and_classes/shard1/tf.nn.moments.md @@ -6,6 +6,9 @@ The mean and variance are calculated by aggregating the contents of `x` across `axes`. If `x` is 1-D and `axes = [0]` this is just the mean and variance of a vector. +Note: for numerical stability, when shift=None, the true mean +would be computed and used as shift. + When using these moments for batch normalization (see `tf.nn.batch_normalization`): @@ -20,8 +23,9 @@ When using these moments for batch normalization (see * `axes`: Array of ints. Axes along which to compute mean and variance. * `shift`: A `Tensor` containing the value by which to shift the data for - numerical stability, or `None` if no shift is to be performed. A shift - close to the true mean provides the most numerically stable results. + numerical stability, or `None` in which case the true mean of the data is + used as shift. A shift close to the true mean provides the most + numerically stable results. * `name`: Name used to scope the operations that compute the moments. * `keep_dims`: produce moments with the same dimensionality as the input. diff --git a/tensorflow/g3doc/api_docs/python/nn.md b/tensorflow/g3doc/api_docs/python/nn.md index fcc73663efd..983f68f8558 100644 --- a/tensorflow/g3doc/api_docs/python/nn.md +++ b/tensorflow/g3doc/api_docs/python/nn.md @@ -1932,6 +1932,9 @@ The mean and variance are calculated by aggregating the contents of `x` across `axes`. If `x` is 1-D and `axes = [0]` this is just the mean and variance of a vector. +Note: for numerical stability, when shift=None, the true mean +would be computed and used as shift. + When using these moments for batch normalization (see `tf.nn.batch_normalization`): @@ -1946,8 +1949,9 @@ When using these moments for batch normalization (see * `axes`: Array of ints. Axes along which to compute mean and variance. * `shift`: A `Tensor` containing the value by which to shift the data for - numerical stability, or `None` if no shift is to be performed. A shift - close to the true mean provides the most numerically stable results. + numerical stability, or `None` in which case the true mean of the data is + used as shift. A shift close to the true mean provides the most + numerically stable results. * `name`: Name used to scope the operations that compute the moments. * `keep_dims`: produce moments with the same dimensionality as the input.