Description of tensor of weights for winograd 2x3 convolution. More...
#include <dnnl_types.h>
Collaboration diagram for dnnl_wino_desc_t:
Public Attributes | |
| dnnl_wino_memory_format_t | wino_format |
| int | r |
| int | alpha |
| int | ic |
| int | oc |
| int | ic_block |
| int | oc_block |
| int | ic2_block |
| int | oc2_block |
| float | adj_scale |
| size_t | size |
Detailed Description
Description of tensor of weights for winograd 2x3 convolution.
Member Data Documentation
◆ adj_scale
| float dnnl_wino_desc_t::adj_scale |
◆ alpha
| int dnnl_wino_desc_t::alpha |
◆ ic
| int dnnl_wino_desc_t::ic |
◆ ic2_block
| int dnnl_wino_desc_t::ic2_block |
◆ ic_block
| int dnnl_wino_desc_t::ic_block |
◆ oc
| int dnnl_wino_desc_t::oc |
◆ oc2_block
| int dnnl_wino_desc_t::oc2_block |
◆ oc_block
| int dnnl_wino_desc_t::oc_block |
◆ r
| int dnnl_wino_desc_t::r |
◆ size
| size_t dnnl_wino_desc_t::size |
◆ wino_format
| dnnl_wino_memory_format_t dnnl_wino_desc_t::wino_format |
The documentation for this struct was generated from the following file:
- /work/mxnet/include/mkldnn/dnnl_types.h
