Intel® oneAPI Deep Neural Network Developer Guide and Reference
A newer version of this document is available. Customers should click here to go to the newest version.
Visible to Intel only — GUID: GUID-08A25DB5-60A6-433B-AB20-E9BC3A59E87C
Visible to Intel only — GUID: GUID-08A25DB5-60A6-433B-AB20-E9BC3A59E87C
enum dnnl_sparse_encoding_t
Overview
Sparse encodings. More…
#include <dnnl_types.h> enum dnnl_sparse_encoding_t { dnnl_sparse_encoding_undef = 0, dnnl_csr, dnnl_packed, dnnl_coo, };
Detailed Documentation
Sparse encodings.
Enum Values
dnnl_sparse_encoding_undef
Undefined sparse encoding kind, used for empty memory descriptors.
dnnl_csr
Compressed Sparse Row (CSR) encoding.
dnnl_packed
An encoding that is used for an opaque storage schema for tensors with unstructured sparsity.
A memory descriptor with the packed encoding cannot be used to create a memory object. It can only be used to create a primitive descriptor to query the actual memory descriptor (similar to the format tag any).
dnnl_coo
Coordinate Sparse Encoding (COO).