ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-04 21:51:28 +00:00
Code Issues Packages Projects Releases Wiki Activity

[CK_TILE] Implement Row/Col quant grouped gemm (#2786)

Browse Source 
* Add cshuffle epilogue test

* add the poc implementation to the epilogue and tests

* refactor cshuffle epilogue

* WIP: adding tensor/tile usage to scale_tile

* fix usage of tile_elementwise_inout

* add gemm_quant_kernel for generalizing gemm quant kernel

* Add problem specific to different quants, add QuantType to Traits

* Add quant_type to quant_kernel template parameters

* Create aq/bq_block_windows and views depending on QuantType

* Use tile windows as inputs in cshuffle epilogue

* Fix some issues in epilogue

* initial new example code for new general gemm quant kernel test

* Fix issues in kernel

* Add verification check for rowcol Quantmode

* use AccDataType instead of AQ in pipeline

* fix aquant preshuffle

* fix formatting

* some cleanup

* remove gemm_aquant_basic.cpp

* remove gemm_aquant_kernel.hpp

* fix tests for the renamed quant kernel

* fix formatting

* clean example files

* fix some merge conflicts

* fix preshufflequant rename issue

* updating

* fix some templates after merging with develop

* fix test preshuffle parameter

* fix formatting

* updating kernels

* change update user

* test username

* update quant_grouped_gemm example

* update example

* Unify bquant kernel to the common quant kernel

* remove bquant kernel also from common header

* fix formatting

* clean up commented code

* update grouped_gemm_quant example

* fix formatting config hpp

* fix merge mistake

* Non-const for movable windows

* fix formatting

* update tileloop pipleline

* Fix grammar in README

Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com>

* Remove #include<bit> and clean up example

* fix strides

* Add some descriptions for move_windows

* fix tensor print bug

* update quant_grouped_gemm example

* remove useless code

* cleanup code

* clean up code & format code

* fix compile & running bug in grouped_gemm example

---------

Co-authored-by: Sami Remes <samremes@amd.com>
Co-authored-by: Mohsen Saffari <mohsen.saffari@amd.com>
Co-authored-by: liyingli <liyingli@amd.com>
Co-authored-by: kyle-256 <Kyle.Zhao@amd.com>
Co-authored-by: spolifroni-amd <Sandra.Polifroni@amd.com>
This commit is contained in:
kyle-256
2025-09-09 01:25:57 +08:00
committed by GitHub
parent 5ff205ca79
commit 4eb415829e
12 changed files with 1225 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
example/ck_tile/17_grouped_gemm/grouped_gemm.cpp
View File

@@ -321,6 +321,36 @@ int run_gemm_example_prec_type(std::string a_layout, std::string b_layout, int a
throw std::runtime_error("Unsupported data layout configuration for A and B tensors!");
}
}
template <template <typename PrecType> typename GemmConfig>
int run_grouped_gemm_example(int argc, char* argv[])
{
auto [result, arg_parser] = create_args(argc, argv);
if(!result)
{
return -1;
}
const std::string a_layout = arg_parser.get_str("a_layout");
const std::string b_layout = arg_parser.get_str("b_layout");
const std::string data_type = arg_parser.get_str("prec");
if(data_type == "fp16")
{
return run_gemm_example_prec_type<GemmConfig<ck_tile::half_t>, ck_tile::half_t>(
a_layout, b_layout, argc, argv);
}
else if(data_type == "fp8")
{
return run_gemm_example_prec_type<GemmConfig<ck_tile::fp8_t>, ck_tile::fp8_t>(
a_layout, b_layout, argc, argv);
}
else
{
throw std::runtime_error("Unsupported data type configuration.");
}
}
int main(int argc, char* argv[])
{
return !run_grouped_gemm_example<GemmConfigComputeV4>(argc, argv);