3 people 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

..

collective

Groupwise scaling along M for FP8 gemm (#2037)

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

device

CUTLASS 3.8 Release (#2059)

2025-01-25 02:44:06 -05:00

kernel

CUTLASS 3.8 Release (#2059)

2025-01-25 02:44:06 -05:00

thread

CUTLASS 3.7 (#2045)

2025-01-18 09:53:07 -05:00

threadblock

CUTLASS 3.7 (#2045)

2025-01-18 09:53:07 -05:00

warp

CUTLASS 3.7 (#2045)

2025-01-18 09:53:07 -05:00

dispatch_policy.hpp

Groupwise scaling along M for FP8 gemm (#2037)

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

gemm_enumerated_types.h

CUTLASS 3.7 (#2045)

2025-01-18 09:53:07 -05:00

gemm.h

CUTLASS 3.7 (#2045)

2025-01-18 09:53:07 -05:00

group_array_problem_shape.hpp

CUTLASS 3.7 (#2045)

2025-01-18 09:53:07 -05:00