A descriptor of a matrix multiplication operation. More...
#include <dnnl_types.h>
Collaboration diagram for dnnl_matmul_desc_t:
Public Attributes | |
| dnnl_primitive_kind_t | primitive_kind |
| The kind of primitive. More... | |
| dnnl_memory_desc_t | src_desc |
| Source memory descriptor. | |
| dnnl_memory_desc_t | weights_desc |
| Weights memory descriptor. | |
| dnnl_memory_desc_t | bias_desc |
| Bias memory descriptor. | |
| dnnl_memory_desc_t | dst_desc |
| Destination memory descriptor. | |
| dnnl_data_type_t | accum_data_type |
| The accumulator data type. Initialized automatically. | |
Detailed Description
A descriptor of a matrix multiplication operation.
2D case: dst[m, n] = src[m, k] * weights[k, n] + bias[m, n]
3D case: dst[mb, m, n] = src[mb, m, k] * weights[mb, k, n] + bias[mb, m, n]
Member Data Documentation
◆ primitive_kind
| dnnl_primitive_kind_t dnnl_matmul_desc_t::primitive_kind |
The kind of primitive.
Used for self-identifying the primitive descriptor. Must be dnnl_matmul.
The documentation for this struct was generated from the following file:
- include/dnnl_types.h