Rationale: the operation does not always sort COO tensors (also used for sparse_tensor.compress for example).
8.2 KiB
8.2 KiB
Rationale: the operation does not always sort COO tensors (also used for sparse_tensor.compress for example).