Adam Osewski a60cf0d0ee Use AccDataType for Output of MFMA instruction.

2024-10-10 09:48:33 +00:00

..

common

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00

epilogue

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00

fmha

[CK_TILE] Simplify the codes in splitkv_combine pipeline (#1549)

2024-10-08 10:44:34 +08:00

gemm

Use AccDataType for Output of MFMA instruction.

2024-10-10 09:48:33 +00:00

image_to_column

[CK_TILE] Image to Column kernel (#1532)

2024-09-27 22:57:38 +02:00

layernorm2d

[Ck tile] Support layernorm one pass (#1512)

2024-10-07 14:25:53 +08:00

reduce/block

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00

welford

layernorm2d forward (#1339)

2024-06-24 08:45:52 +08:00

common.hpp

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00

epilogue.hpp

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00

fmha.hpp

[CK_TILE] Add PagedAttention kernels (#1387)

2024-08-28 20:50:43 +08:00

gemm.hpp

Get hot loop and TailNum information before lunching kernel.

2024-10-01 14:40:03 +00:00

image_to_column.hpp

[CK_TILE] add missing vector header (#1537)

2024-10-01 07:58:20 -07:00

layernorm2d.hpp

layernorm2d forward (#1339)

2024-06-24 08:45:52 +08:00

reduce.hpp

introducing ck_tile! (#1216)

2024-04-15 19:27:12 -05:00

welford.hpp

layernorm2d forward (#1339)

2024-06-24 08:45:52 +08:00