Intel(R) Math Kernel Library for Deep Neural Networks (Intel(R) MKL-DNN)  1.0.4
Performance library for Deep Learning
Classes | List of all members
mkldnn::batch_normalization_backward Struct Reference

Batch normalization backward propagation. More...

#include <mkldnn.hpp>

Inheritance diagram for mkldnn::batch_normalization_backward:
Inheritance graph
[legend]
Collaboration diagram for mkldnn::batch_normalization_backward:
Collaboration graph
[legend]

Classes

struct  desc
 Descriptor for batch normalization backward propagation. More...
 
struct  primitive_desc
 Primitive descriptor for batch normalization backward propagation. More...
 

Additional Inherited Members

- Public Types inherited from mkldnn::primitive
enum  kind {
  kind::undef = mkldnn_undefined_primitive, kind::reorder = mkldnn_reorder, kind::shuffle = mkldnn_shuffle, kind::concat = mkldnn_concat,
  kind::sum = mkldnn_sum, kind::convolution = mkldnn_convolution, kind::deconvolution = mkldnn_deconvolution, kind::eltwise = mkldnn_eltwise,
  kind::softmax = mkldnn_softmax, kind::pooling = mkldnn_pooling, kind::lrn = mkldnn_lrn, kind::batch_normalization = mkldnn_batch_normalization,
  kind::inner_product = mkldnn_inner_product, kind::rnn = mkldnn_rnn
}
 Kinds of primitives. More...
 
- Public Member Functions inherited from mkldnn::primitive
const_mkldnn_primitive_desc_t get_primitive_desc () const
 Returns the descriptor of the underlying C API primitive.
 
- Public Member Functions inherited from mkldnn::handle< mkldnn_primitive_t >
 handle (mkldnn_primitive_t t, bool weak=false)
 Constructs a C handle wrapper. More...
 
 handle ()
 Empty constructor. More...
 
void reset (mkldnn_primitive_t t, bool weak=false)
 Resets the value of a C handle. More...
 
mkldnn_primitive_t get (bool allow_emtpy=false) const
 Returns the value of the underlying C handle.
 

Detailed Description

Batch normalization backward propagation.

Implements descriptor, primitive descriptor, and primitive.


The documentation for this struct was generated from the following file: