Closed paul-tqh-nguyen closed 3 years ago
In several places throughout our code base (e.g. in tersify_mlir) and documentation, we have two names for sparse vector encodings, #SparseVec64 and #CV64.
tersify_mlir
#SparseVec64
#CV64
#SparseVec64 = #sparse_tensor.encoding<{ dimLevelType = [ "compressed" ], pointerBitWidth = 64, indexBitWidth = 64 }>
We should be consistent and pick one. I vote for #CV64.
+1 for #CV64
https://github.com/metagraph-dev/mlir-graphblas/pull/191 replaces #SparseVec64 with #CV64.
In several places throughout our code base (e.g. in
tersify_mlir
) and documentation, we have two names for sparse vector encodings,#SparseVec64
and#CV64
.We should be consistent and pick one. I vote for
#CV64
.