oneAPI Deep Neural Network Library (oneDNN)
Performance library for Deep Learning
1.96.0
dnnl::reduction::desc Struct Reference

Descriptor for reduction. More...

#include <dnnl.hpp>

Collaboration diagram for dnnl::reduction::desc:

Public Member Functions

 desc ()=default
 Default constructor. Produces an empty object.
 
 desc (algorithm aalgorithm, const memory::desc &src_desc, const memory::desc &dst_desc, float p, float eps)
 Constructs a descriptor for a reduction primitive using algorithm specific parameters, source and destination memory descriptors. More...
 

Detailed Description

Descriptor for reduction.

Examples:
reduction.cpp.

Constructor & Destructor Documentation

◆ desc()

dnnl::reduction::desc::desc ( algorithm  aalgorithm,
const memory::desc src_desc,
const memory::desc dst_desc,
float  p,
float  eps 
)
inline

Constructs a descriptor for a reduction primitive using algorithm specific parameters, source and destination memory descriptors.

Note
Destination memory descriptor may be initialized with dnnl::memory::format_tag::any value of format_tag.
Parameters
aalgorithmreduction algorithm kind. Possible values: dnnl_reduction_max, dnnl_reduction_min, dnnl_reduction_sum, dnnl_reduction_mul, dnnl_reduction_mean, dnnl_reduction_norm_lp_max, dnnl_reduction_norm_lp_sum, dnnl_reduction_norm_lp_power_p_max, dnnl_reduction_norm_lp_power_p_sum.
palgorithm specific parameter.
epsalgorithm specific parameter.
src_descSource memory descriptor.
dst_descDestination memory descriptor.

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