Public Attributes | List of all members
dnnl_pooling_desc_t Struct Reference

A descriptor of a pooling operation. More...

#include <dnnl_types.h>

Collaboration diagram for dnnl_pooling_desc_t:
Collaboration graph

Public Attributes

dnnl_primitive_kind_t primitive_kind
dnnl_prop_kind_t prop_kind
dnnl_alg_kind_t alg_kind
dnnl_memory_desc_t src_desc
 Source memory descriptor. More...
dnnl_memory_desc_t diff_src_desc
 Source gradient memory descriptor. More...
dnnl_memory_desc_t dst_desc
 Destination memory descriptor. More...
dnnl_memory_desc_t diff_dst_desc
 Destination gradient memory descriptor. More...
dnnl_dims_t strides
 Pooling kernel strides for spatial dimensions. More...
dnnl_dims_t kernel
 Pooling kernel spatial dimensions. More...
dnnl_dims_t padding [2]
dnnl_data_type_t accum_data_type
 The accumulator data type. Initialized automatically. More...

Detailed Description

A descriptor of a pooling operation.

Member Data Documentation

dnnl_data_type_t dnnl_pooling_desc_t::accum_data_type

The accumulator data type. Initialized automatically.

dnnl_alg_kind_t dnnl_pooling_desc_t::alg_kind

The kind of pooling algorithm. Possible values: dnnl_pooling_max, dnnl_pooling_avg_include_padding, and dnnl_pooling_avg_exclude_padding.

dnnl_memory_desc_t dnnl_pooling_desc_t::diff_dst_desc

Destination gradient memory descriptor.

dnnl_memory_desc_t dnnl_pooling_desc_t::diff_src_desc

Source gradient memory descriptor.

dnnl_memory_desc_t dnnl_pooling_desc_t::dst_desc

Destination memory descriptor.

dnnl_dims_t dnnl_pooling_desc_t::kernel

Pooling kernel spatial dimensions.

dnnl_dims_t dnnl_pooling_desc_t::padding[2]

Padding in each spatial dimension. padding[0] is a padding in the beginning (padding_l), padding[1] is a padding in the end (padding_r).

dnnl_primitive_kind_t dnnl_pooling_desc_t::primitive_kind

The kind of primitive. Used for self-identifying the primitive descriptor. Must be dnnl_pooling.

dnnl_prop_kind_t dnnl_pooling_desc_t::prop_kind

The kind of propagation. Possible values: dnnl_forward_training, dnnl_forward_inference, dnnl_backward, and dnnl_backward_data.

dnnl_memory_desc_t dnnl_pooling_desc_t::src_desc

Source memory descriptor.

dnnl_dims_t dnnl_pooling_desc_t::strides

Pooling kernel strides for spatial dimensions.

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