Deep Neural Network Library (DNNL)  1.1.3
Performance library for Deep Learning
Functions

A primitive to concatenate data by arbitrary dimension. More...

Functions

dnnl_status_t DNNL_API dnnl_concat_primitive_desc_create (dnnl_primitive_desc_t *concat_primitive_desc, const dnnl_memory_desc_t *dst_md, int n, int concat_dimension, const dnnl_memory_desc_t *src_mds, const_dnnl_primitive_attr_t attr, dnnl_engine_t engine)
 Creates out-of-place concat_primitive_desc for concatenation of n inputs by concat_dimension with resulting output_desc memory descriptor. More...
 

Detailed Description

A primitive to concatenate data by arbitrary dimension.

See also
Concat in developer guide
Concat in C++ API

Function Documentation

◆ dnnl_concat_primitive_desc_create()

dnnl_status_t DNNL_API dnnl_concat_primitive_desc_create ( dnnl_primitive_desc_t concat_primitive_desc,
const dnnl_memory_desc_t dst_md,
int  n,
int  concat_dimension,
const dnnl_memory_desc_t src_mds,
const_dnnl_primitive_attr_t  attr,
dnnl_engine_t  engine 
)

Creates out-of-place concat_primitive_desc for concatenation of n inputs by concat_dimension with resulting output_desc memory descriptor.

output_desc can be NULL or specified with the dnnl_format_kind_any format kind – in this case, the appropriate memory format would be chosen automatically.

Inputs:

Outputs: