Liang 3c28697b9f Groupwise scaling along M for FP8 gemm (#2037) ...

* FP8 groupwise scaling along M

* small updates

---------

Co-authored-by: zl <zl@deepseek.com>
Co-authored-by: Haicheng Wu <haichengw@nvidia.com>

2025-01-31 13:51:28 -05:00

..

cute

fix cuda 12.6 issues (#2066)

2025-01-28 17:28:29 -05:00

cutlass

Groupwise scaling along M for FP8 gemm (#2037)

2025-01-31 13:51:28 -05:00