mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-01 20:21:23 +00:00

Files

rocking abae2afc72 support max3 in smoothquant and add+ rmsnorm + rdquant (#1654 )

* Fix cmake example build

* Support max3 in smoothquant one pass

* support max3 in two pass

* support max3 in add_rmsnorm_rdquant

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

instances

[Ck_tile] smoothquant (#1617 )

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

script

[Ck_tile] smoothquant (#1617 )

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

CMakeLists.txt

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

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

example_smoothquant.cpp

[Ck_tile] smoothquant (#1617 )

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

README.md

[Ck_tile] smoothquant (#1617 )

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

smoothquant.cpp

[Ck_tile] smoothquant (#1617 )

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

smoothquant.hpp

[Ck_tile] smoothquant (#1617 )

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

README.md

smoothquant

This folder contains example for smoothquant using ck_tile tile-programming implementation.

build

# in the root of ck_tile
mkdir build && cd build
sh ../script/cmake-ck-dev.sh  ../ <arch>  # you can replace this <arch> to gfx90a, gfx942...
make tile_smoothquant -j

This will result in an executable build/bin/tile_smoothquant

cmdline

args:
          -m    m dimension (default:3328)
          -n    m dimension (default:4096)
          -v    cpu validation or not (default:1)
       -prec    precision (default:fp16)