jakpiase e7b6286441 Add interwave scheduler for gemm mem pipeline (#1647) ...

* add interwave scheduler for gemm mem pipeline

* Fix merge artifacts.

* Refactor unit tests.

* Switch to interwave scheduler for mem example

---------

Co-authored-by: Adam Osewski <19374865+aosewski@users.noreply.github.com>
Co-authored-by: Adam Osewski <Adam.Osewski@amd.com>

2024-11-27 18:25:07 +01:00

..

add_rmsnorm2d_rdquant

support max3 in smoothquant and add+ rmsnorm + rdquant (#1654)

2024-11-27 05:01:15 +08:00

common

[Ck tile] layernorm2d fwd optimize (#1637)

2024-11-08 12:28:23 +08:00

elementwise

[CK_TILE] fused-moe first version (#1634)

2024-11-26 11:14:56 +08:00

epilogue

[CK_TILE] layernorm have more accurate residual (#1623)

2024-11-02 13:30:16 +08:00

flatmm/block

[CK_TILE] fused-moe first version (#1634)

2024-11-26 11:14:56 +08:00

fmha

[CK_TILE] Fix incorrect computation of group mode PagedAttention (#1688)

2024-11-26 20:37:54 +08:00

fused_moe

[CK_TILE] fused-moe first version (#1634)

2024-11-26 11:14:56 +08:00

gemm

Add interwave scheduler for gemm mem pipeline (#1647)

2024-11-27 18:25:07 +01:00

image_to_column

[CK_TILE] Image to Column kernel (#1532)

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

layernorm2d

[Ck_tile] hot fix, fix rpcf param setting err (#1657)

2024-11-14 14:06:36 +08:00

permute

Remove virtual destructors from unary ops (#1610)

2024-10-30 17:42:50 +01:00

reduce/block

[Ck_tile] smoothquant (#1617)

2024-11-01 13:51:56 +08:00

rmsnorm2d

[Ck_tile] smoothquant (#1617)

2024-11-01 13:51:56 +08:00

smoothquant

support max3 in smoothquant and add+ rmsnorm + rdquant (#1654)

2024-11-27 05:01:15 +08:00

softmax/block

topk_softmax (#1592)

2024-10-26 23:52:49 +08:00

topk/block

topk_softmax (#1592)

2024-10-26 23:52:49 +08:00

topk_softmax

topk_softmax (#1592)

2024-10-26 23:52:49 +08:00

welford

[Ck_tile] hot fix, fix rpcf param setting err (#1657)

2024-11-14 14:06:36 +08:00

add_rmsnorm2d_rdquant.hpp

[Ck_tile] smoothquant (#1617)

2024-11-01 13:51:56 +08:00

common.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

elementwise.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

epilogue.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

flatmm.hpp

[CK_TILE] fused-moe first version (#1634)

2024-11-26 11:14:56 +08:00

fmha.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

fused_moe.hpp

[CK_TILE] fused-moe first version (#1634)

2024-11-26 11:14:56 +08:00

gemm.hpp

CK-Tile first draft of universal block gemm with interwave & intrawave scheduler (#1676)

2024-11-26 08:45:14 +01:00

image_to_column.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

layernorm2d.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

permute.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

reduce.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

rmsnorm2d.hpp

[Ck_tile] smoothquant (#1617)

2024-11-01 13:51:56 +08:00

smoothquant.hpp

[CK_TILE]Moe update index (#1672)

2024-11-25 13:12:35 +08:00

softmax.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

topk_softmax.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

topk.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00

welford.hpp

[CK_TILE] layernorm support fused-quant/fused-add (#1604)

2024-10-31 14:54:53 +08:00