mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-05-14 02:02:46 +00:00
bd01797525
* use 'sweep once' softmax kernel where applicable
* threadwise copy's dst buffer can specify invalid element value
* add int8 in/out float compute softmax support
give a bit of leeway for int absolute tolerance as there's a single data point of all test cases showing off-by-1 error
* format
* softmax inherits DeviceNormalization
* softmax profiler stub
* tighten up reference softmax interface
* example prints tensor dimension
* add fp32 to softmax profiler
* rename header
* hook with ckProfiler
* format
* resolve merge conflict
* resolve merge conflicts
* update normalization profiler help string
* resolve conflict
* typo
* remove residual
* softmax profiler: address feedback
* test for mixed precision input/output
* fully qualify ck::math::isnan
* add comment for device normalization interface
* revise wording
* constness for alpha/beta scaler pointer
[ROCm/composable_kernel commit: 93c99f3d87]
35 lines
2.0 KiB
C++
35 lines
2.0 KiB
C++
// SPDX-License-Identifier: MIT
|
|
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
|
|
|
|
#include "gtest/gtest.h"
|
|
#include "test_softmax_util.hpp"
|
|
|
|
template <ck::index_t N>
|
|
using I = ck::Number<N>;
|
|
|
|
template <typename Tuple>
|
|
class TestSoftmaxFP32 : 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<float, float, ck::half_t, I<3>, I<2>, I<256>, I<1>, I<256>, I<1>, I<8>, I<1>, I<4>, I<8>>, // mixed precision
|
|
std::tuple<float, float, float, I<3>, I<1>, I<256>, I<8>, I<32>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<1>, I<256>, I<4>, I<64>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<1>, I<256>, I<2>, I<128>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<1>, I<256>, I<1>, I<256>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<1>, I<256>, I<1>, I<256>, I<1>, I<8>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<1>, I<256>, I<1>, I<256>, I<1>, I<16>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<2>, I<256>, I<8>, I<32>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<2>, I<256>, I<4>, I<64>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<2>, I<256>, I<2>, I<128>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<2>, I<256>, I<1>, I<256>, I<1>, I<4>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<2>, I<256>, I<1>, I<256>, I<1>, I<8>, I<1>, I<4>, I<4>>,
|
|
std::tuple<float, float, float, I<3>, I<2>, I<256>, I<1>, I<256>, I<1>, I<16>, I<1>, I<4>, I<4>>
|
|
>;
|
|
// clang-format on
|
|
TYPED_TEST_SUITE(TestSoftmaxFP32, KernelTypes);
|
|
TYPED_TEST(TestSoftmaxFP32, Test_FP32) { this->Run(); }
|