[CPU] Add FP8 W8A16 linear support (#41186)

Signed-off-by: yuwenzho <yuwen.zhou@intel.com>
This commit is contained in:
Yuwen Zhou
2026-05-06 07:05:27 +00:00
committed by GitHub
parent b53c507bc9
commit 809b98e5b7
9 changed files with 331 additions and 3 deletions
+1 -1
View File
@@ -447,7 +447,7 @@ INSTANTIATE_TINYGEMM_TEMPLATE(at::BFloat16);
INSTANTIATE_TINYGEMM_TEMPLATE(at::Half);
at::Tensor fp8_scaled_mm_cpu(at::Tensor& mat1, at::Tensor& mat2, at::Tensor& scales2,
std::vector<int64_t> block_size, std::optional<at::Tensor>& bias,
std::vector<int64_t> block_size, const std::optional<at::Tensor>& bias,
at::ScalarType out_dtype, bool is_vnni) {
RECORD_FUNCTION("sgl-kernel::fp8_scaled_mm_cpu", std::vector<c10::IValue>({mat1, mat2, scales2, block_size, bias}));