mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-14 18:17:44 +00:00
08b531ae32
* initial stub for standalone softmax
* start device_softmax_mk_to_mk as a wrapper to device_reduce_mk_to_m
* host softmax validates
* compiles; to implement beta scaling
* use NaN trick to efficiently ignore OOB values during sum of exponentials
* freeload device_reduce's utility functions
* clean up interface
* adding prior value (beta scaling)
* remove restriction related to perf considerations
* apply clang-format
* clean; disable diagnostics
* resolve conflicts
* add exp wrapper
* honor HostTensorDesc interface; allow implicit cast from different vector<T> type
* test softmax for fp16/fp32
* update readme
* amend commit NaN trick
* remove redundant param added during development
* format
* replace ScalarDataType with AccDataType
* separate out test programs by precision type
* move softmax sample code to its own folder
* format
* keep up with recent changes in reduction API
* remove extra header
[ROCm/composable_kernel commit: 15c89e81f0]
27 lines
1.4 KiB
C++
27 lines
1.4 KiB
C++
#include "gtest/gtest.h"
|
|
#include "test_softmax_util.hpp"
|
|
|
|
template <ck::index_t N>
|
|
using I = ck::Number<N>;
|
|
|
|
template <typename Tuple>
|
|
class TestSoftmaxFP16 : public ck::TestSoftmax<Tuple>
|
|
{
|
|
};
|
|
|
|
// clang-format off
|
|
using KernelTypes = ::testing::Types<
|
|
// InDataType, AccDataType, OutDataType, Rank, NumReduceDim, BlockSize, MThreadClusterSize, KThreadClusterSize, MThreadSliceSize, KThreadSliceSize, InSrcVectorDim, InSrcVectorSize, OutDstVectorSize>
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<1>, I<256>, I<8>, I<32>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<1>, I<256>, I<4>, I<64>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<1>, I<256>, I<2>, I<128>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<1>, I<256>, I<1>, I<256>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<2>, I<256>, I<8>, I<32>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<2>, I<256>, I<4>, I<64>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<2>, I<256>, I<2>, I<128>, I<1>, I<8>, I<1>, I<8>, I<8>>,
|
|
std::tuple<ck::half_t, float, ck::half_t, I<3>, I<2>, I<256>, I<1>, I<256>, I<1>, I<8>, I<1>, I<8>, I<8>>
|
|
>;
|
|
// clang-format on
|
|
TYPED_TEST_SUITE(TestSoftmaxFP16, KernelTypes);
|
|
TYPED_TEST(TestSoftmaxFP16, Test_FP16) { this->Run(); }
|