ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-17 11:30:02 +00:00
Code Issues Packages Projects Releases Wiki Activity

Softmax unit-test reduction across all and non innermost dims cases. (#406)

Browse Source 
* Add reduction across all dims cases.

* host softmax: handle all reduce

* Test cases when reduced dim is not innermost axis.

* Fix syntax.

* Test non innermost dim for fp32 and int8

* Group test suites wrt NumReduceDim.

* Additionally test failing cases.

* Throw error when Rank or NumReduceDims doesn't match arguments.

* Check reducedDims has correct values

* Move don't reuse DeviceReduceMultiblock IsSupportedArgument method.
Instead implement own. (in fact just get rid of one check to enable
reduction across inner dimensions).

* Reorganize unit tests to better cover use scenarios.

* Test input validation
* Test reduction of inner dimensions with custom op instances.

* Refactor fp32 and int8 unit tests.

* Fix FP32 instance template parameters.

* Add more instances.

* Instances with InSrcVectorDim=0.

* Do not initialize and copy data when arg not supported.

* ckProfiler Softmax use instance factory.

* Refactor device softmax IsSupported.

* Additionally add non-polymorphic api functions

* Split softmax instances into multiple files.

* Fix profiler.

* Reorganize tests to reuse profiler and cover edge cases.

* Clang-format

* I8 Softmax instances along with UT.

* Reuse type alias definitions from instance factory header.

* Clean included headers

* Fix variable names.

* Add missing checks in Argument constructor.

Co-authored-by: Adam Osewski <aosewski@amd.com>
Co-authored-by: Anthony Chang <ac.chang@outlook.com>

[ROCm/composable_kernel commit: 6d8614ee50]
This commit is contained in:
Adam Osewski
2022-11-02 23:46:08 +01:00
committed by GitHub
parent 9c3ea1fcc5
commit efd02ae56c
71 changed files with 1872 additions and 468 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
profiler/src/profiler.cpp
View File

@@ -20,7 +20,7 @@ int profile_conv_fwd_bias_relu_add(int, char*[]);
int profile_conv_bwd_data(int, char*[]);
int profile_conv_bwd_weight(int, char*[]);
int profile_grouped_conv_fwd(int, char*[]);
int profile_normalization(int, char*[]);
int profile_softmax(int, char*[]);
int profile_layernorm(int, char*[]);
int profile_groupnorm(int, char*[]);
int profile_reduce(int, char*[]);
@@ -45,6 +45,7 @@ static void print_helper_message()
" conv_bwd_data: Convolution Backward Data\n"
" conv_bwd_weight: Convolution Backward Weight\n"
" grouped_conv_fwd: Grouped Convolution Forward\n"
" softmax: Softmax\n"
" reduce: Reduce\n");
// clang-format on
}
@@ -129,9 +130,9 @@ int main(int argc, char* argv[])
{
return profile_reduce(argc, argv);
}
else if(strcmp(argv[1], "batchnorm") == 0 || strcmp(argv[1], "softmax") == 0)
else if(strcmp(argv[1], "softmax") == 0)
{
return profile_normalization(argc, argv);
return profile_softmax(argc, argv);
}
else if(strcmp(argv[1], "layernorm") == 0)
{