tensorflow:: ops:: FusedBatchNormGrad:: Attrs


   #include <nn_ops.h>

Optional attribute setters for FusedBatchNormGrad .

Summary

Public functions
`DataFormat (StringPiece x)`	`TF_MUST_USE_RESULT Attrs` The data format for y_backprop, x, x_backprop.
`Epsilon (float x)`	`TF_MUST_USE_RESULT Attrs` A small float number added to the variance of x.
`IsTraining (bool x)`	`TF_MUST_USE_RESULT Attrs` A bool value to indicate the operation is for training (default) or inference.

StringPiece tensorflow::ops::FusedBatchNormGrad::Attrs::data_format_ = "NHWC"

float tensorflow::ops::FusedBatchNormGrad::Attrs::epsilon_ = 0.0001f

bool tensorflow::ops::FusedBatchNormGrad::Attrs::is_training_ = true

TF_MUST_USE_RESULT Attrs tensorflow::ops::FusedBatchNormGrad::Attrs::DataFormat(
  StringPiece x
)

The data format for y_backprop, x, x_backprop.

Either "NHWC" (default) or "NCHW".

Defaults to "NHWC"

TF_MUST_USE_RESULT Attrs tensorflow::ops::FusedBatchNormGrad::Attrs::Epsilon(
  float x
)

A small float number added to the variance of x.

Defaults to 0.0001

TF_MUST_USE_RESULT Attrs tensorflow::ops::FusedBatchNormGrad::Attrs::IsTraining(
  bool x
)

A bool value to indicate the operation is for training (default) or inference.

Defaults to true