[Kernel][Bugfix] Refactor and Fix CUTLASS 2:4 Sparse Kernels (#13198)

Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
This commit is contained in:
Tyler Michael Smith
2025-02-14 00:01:14 +00:00
committed by GitHub
parent 2344192a55
commit c1e37bf71b
16 changed files with 576 additions and 473 deletions
+1 -2
View File
@@ -176,8 +176,7 @@ void cutlass_scaled_sparse_mm(torch::Tensor& out, torch::Tensor const& a,
torch::Tensor const& b_scales,
std::optional<torch::Tensor> const& bias);
bool cutlass_sparse_compress_entry(torch::Tensor& a_compressed,
torch::Tensor& e, torch::Tensor const& a);
std::vector<torch::Tensor> cutlass_sparse_compress(torch::Tensor const& a);
#endif
void static_scaled_int8_quant(torch::Tensor& out, torch::Tensor const& input,