Illia Silin 64dc32a54b Put back the split-k gemm code. (#684) ...

* simplify karg in device/grid split-k op

* fix mk_kn_mn instances

* add more instances

* use name from tensor layout

---------

Co-authored-by: carlushuang <carlus.huang@amd.com>

[ROCm/composable_kernel commit: 903cd19ce3]

2023-04-21 19:37:00 -05:00

..

host_utility

Clean up kernel launch output (#569)

2023-02-15 12:07:21 -06:00

problem_transform

add license in file (#303)

2022-06-24 23:32:43 -05:00

tensor

add license in file (#303)

2022-06-24 23:32:43 -05:00

tensor_description

Input/output permutation for fused attention (#460)

2022-10-27 14:58:20 -06:00

tensor_operation

Put back the split-k gemm code. (#684)

2023-04-21 19:37:00 -05:00

utility

Groupnorm + swish external api (#668)

2023-04-10 08:02:17 -05:00

ck.hpp

add a marco to turn on/off denorm fix (off by default) (#673)

2023-04-11 07:44:43 -05:00

stream_config.hpp

disable print for group conv multiple D (#421)

2022-09-16 09:46:32 -05:00