ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-20 06:49:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add dynamic elementwise op (#1426)

Browse Source 
* Add dynamic elementwise op

Co-authored-by: ThruptiRajLakshmanaGowda <thruptiraj.lakshmanagowda@amd.com>

* CI issues fix

* Custom parameter value for dynamic functions - Comments addressed

---------

Co-authored-by: ThruptiRajLakshmanaGowda <thruptiraj.lakshmanagowda@amd.com>
Co-authored-by: ThruptiRajLakshmanaGowda <tlakshma@amd.com>
This commit is contained in:
Bartłomiej Kocot
2024-10-26 15:22:37 +02:00
committed by GitHub
parent 54f0e6f4bb
commit 31bf253aeb
32 changed files with 2194 additions and 170 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
example/62_convnd_activ/CMakeLists.txt
View File

@@ -6,6 +6,7 @@ add_subdirectory(convscale_add)
add_subdirectory(convscale_reduce)
add_subdirectory(multi_AB)
add_subdirectory(unary)
add_subdirectory(dynamic_unary)
add_custom_target(example_convnd_activ_xdl)
# ScaleAdd ScaleAdd Relu

45
example/62_convnd_activ/dynamic_unary/CMakeLists.txt Normal file
View File

@@ -0,0 +1,45 @@
list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
set(target 0)
foreach(gpu IN LISTS GPU_TARGETS)
if(gpu IN_LIST gpu_list AND target EQUAL 0)
add_custom_target(example_convnd_activ_dynamic_unary_xdl)
# Sigmoid
add_example_executable(example_convnd_fwd_xdl_dynamic_sigmoid_fp16 convnd_fwd_xdl_dynamic_sigmoid_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_sigmoid_fp16)
# Tanh
add_example_executable(example_convnd_fwd_xdl_dynamic_tanh_fp16 convnd_fwd_xdl_dynamic_tanh_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_tanh_fp16)
# Relu
add_example_executable(example_convnd_fwd_xdl_dynamic_relu_fp16 convnd_fwd_xdl_dynamic_relu_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_relu_fp16)
# SoftRelu
add_example_executable(example_convnd_fwd_xdl_dynamic_softrelu_fp16 convnd_fwd_xdl_dynamic_softrelu_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_softrelu_fp16)
# Abs
add_example_executable(example_convnd_fwd_xdl_dynamic_abs_fp16 convnd_fwd_xdl_dynamic_abs_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_abs_fp16)
# Pow
add_example_executable(example_convnd_fwd_xdl_dynamic_pow_fp16 convnd_fwd_xdl_dynamic_pow_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_pow_fp16)
# Clipped Relu
add_example_executable(example_convnd_fwd_xdl_dynamic_clippedrelu_fp16 convnd_fwd_xdl_dynamic_clippedrelu_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_clippedrelu_fp16)
# Leaky Relu
add_example_executable(example_convnd_fwd_xdl_dynamic_leakyrelu_fp16 convnd_fwd_xdl_dynamic_leakyrelu_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_leakyrelu_fp16)
# Elu
add_example_executable(example_convnd_fwd_xdl_dynamic_elu_fp16 convnd_fwd_xdl_dynamic_elu_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_elu_fp16)
# Swish
add_example_executable(example_convnd_fwd_xdl_dynamic_swish_fp16 convnd_fwd_xdl_dynamic_swish_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_swish_fp16)
# PassThrough
add_example_executable(example_convnd_fwd_xdl_dynamic_passthrough_fp16 convnd_fwd_xdl_dynamic_passthrough_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_passthrough_fp16)
# Logistic
add_example_executable(example_convnd_fwd_xdl_dynamic_logistic_fp16 convnd_fwd_xdl_dynamic_logistic_fp16.cpp)
add_example_dependencies(example_convnd_activ_dynamic_unary_xdl example_convnd_fwd_xdl_dynamic_logistic_fp16)
set(target 1)
endif()
endforeach()

238
example/62_convnd_activ/dynamic_unary/convnd_fwd_activ_dynamic_unary_common.hpp Normal file
View File

@@ -0,0 +1,238 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
#include <cstdlib>
#include <iostream>
#include <numeric>
#include <type_traits>
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
#include "ck/tensor_operation/gpu/device/impl/device_grouped_conv_fwd_multiple_abd_xdl_cshuffle.hpp"
#include "ck/library/utility/algorithm.hpp"
#include "ck/library/utility/check_err.hpp"
#include "ck/library/utility/device_memory.hpp"
#include "ck/library/utility/host_tensor.hpp"
#include "ck/library/utility/host_tensor_generator.hpp"
#include "ck/library/utility/convolution_parameter.hpp"
#include "ck/library/utility/convolution_host_tensor_descriptor_helper.hpp"
#include "ck/library/reference_tensor_operation/cpu/reference_conv_fwd.hpp"
#include "ck/library/utility/convolution_host_tensor_descriptor_helper.hpp"
constexpr ck::index_t NDimSpatial = 3;
using InDataType = ck::half_t;
using WeiDataType = ck::half_t;
using AccDataType = float;
using CShuffleDataType = ck::half_t;
using OutDataType = ck::half_t;
template <ck::index_t... Is>
using S = ck::Sequence<Is...>;
using InLayout = ck::tensor_layout::convolution::GNDHWC;
using WeiLayout = ck::tensor_layout::convolution::GKZYXC;
using OutLayout = ck::tensor_layout::convolution::GNDHWK;
using InElementOp = ck::tensor_operation::element_wise::PassThrough;
using WeiElementOp = ck::tensor_operation::element_wise::PassThrough;
using DynamicElementOp = ck::tensor_operation::element_wise::DynamicUnaryOp;
static constexpr auto ConvSpec =
ck::tensor_operation::device::ConvolutionForwardSpecialization::Default;
static constexpr auto GemmSpec = ck::tensor_operation::device::GemmSpecialization::MNKPadding;
using DeviceGroupedConvNDActivInstance =
ck::tensor_operation::device::DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<
NDimSpatial,
InLayout,
WeiLayout,
ck::Tuple<>,
OutLayout,
InDataType,
WeiDataType,
AccDataType,
CShuffleDataType,
ck::Tuple<>,
OutDataType,
InElementOp,
WeiElementOp,
DynamicElementOp,
ConvSpec, // ConvForwardSpecialization
GemmSpec, // GemmSpecialization
1, //
256, // BlockSize
128, // MPerBlock
256, // NPerBlock
32, // KPerBlock
8, // AK1
8, // BK1
32, // MPerXdl
32, // NPerXdl
2, // MXdlPerWave
4, // NXdlPerWave
S<4, 64, 1>, // ABlockTransferThreadClusterLengths_AK0_M_AK1
S<1, 0, 2>, // ABlockTransferThreadClusterArrangeOrder
S<1, 0, 2>, // ABlockTransferSrcAccessOrder
2, // ABlockTransferSrcVectorDim
8, // ABlockTransferSrcScalarPerVector
8, // ABlockTransferDstScalarPerVector_AK1
1, // ABlockLdsExtraM
S<4, 64, 1>, // BBlockTransferThreadClusterLengths_BK0_N_BK1
S<1, 0, 2>, // BBlockTransferThreadClusterArrangeOrder
S<1, 0, 2>, // BBlockTransferSrcAccessOrder
2, // BBlockTransferSrcVectorDim
8, // BBlockTransferSrcScalarPerVector
8, // BBlockTransferDstScalarPerVector_BK1
1, // BBlockLdsExtraN
1,
1,
S<1, 32, 1, 8>,
8>;
template <ck::index_t NDimSpatial,
typename InDataType,
typename WeiDataType,
typename OutDataType,
typename InElementOp,
typename WeiElementOp,
typename OutElementOp,
typename DeviceConvNDFwdInstance>
bool run_grouped_conv(bool do_verification,
int init_method,
bool time_kernel,
const ck::utils::conv::ConvParam& conv_param,
const HostTensorDescriptor& in_g_n_c_wis_desc,
const HostTensorDescriptor& wei_g_k_c_xs_desc,
const HostTensorDescriptor& out_g_n_k_wos_desc,
const InElementOp& in_element_op,
const WeiElementOp& wei_element_op,
const OutElementOp& out_element_op)
{
Tensor<InDataType> in(in_g_n_c_wis_desc);
Tensor<WeiDataType> wei(wei_g_k_c_xs_desc);
Tensor<OutDataType> out_host(out_g_n_k_wos_desc);
Tensor<OutDataType> out_device(out_g_n_k_wos_desc);
std::cout << "in: " << in.mDesc << std::endl;
std::cout << "wei: " << wei.mDesc << std::endl;
std::cout << "out: " << out_host.mDesc << std::endl;
switch(init_method)
{
case 0: break;
case 1:
in.GenerateTensorValue(GeneratorTensor_2<InDataType>{-2, 2});
wei.GenerateTensorValue(GeneratorTensor_2<WeiDataType>{-2, 2});
break;
default:
in.GenerateTensorValue(GeneratorTensor_3<InDataType>{-1.0, 1.0});
wei.GenerateTensorValue(GeneratorTensor_3<WeiDataType>{-0.05, 0.05});
}
DeviceMem in_device_buf(sizeof(InDataType) * in.mDesc.GetElementSpaceSize());
DeviceMem wei_device_buf(sizeof(WeiDataType) * wei.mDesc.GetElementSpaceSize());
DeviceMem out_device_buf(sizeof(OutDataType) * out_device.mDesc.GetElementSpaceSize());
in_device_buf.ToDevice(in.mData.data());
wei_device_buf.ToDevice(wei.mData.data());
std::array<ck::index_t, NDimSpatial + 3> a_g_n_c_wis_lengths{};
std::array<ck::index_t, NDimSpatial + 3> a_g_n_c_wis_strides{};
std::array<ck::index_t, NDimSpatial + 3> b_g_k_c_xs_lengths{};
std::array<ck::index_t, NDimSpatial + 3> b_g_k_c_xs_strides{};
std::array<ck::index_t, NDimSpatial + 3> e_g_n_k_wos_lengths{};
std::array<ck::index_t, NDimSpatial + 3> e_g_n_k_wos_strides{};
std::array<ck::index_t, NDimSpatial> conv_filter_strides{};
std::array<ck::index_t, NDimSpatial> conv_filter_dilations{};
std::array<ck::index_t, NDimSpatial> input_left_pads{};
std::array<ck::index_t, NDimSpatial> input_right_pads{};
auto copy = [](const auto& x, auto& y) { ck::ranges::copy(x, y.begin()); };
copy(in_g_n_c_wis_desc.GetLengths(), a_g_n_c_wis_lengths);
copy(in_g_n_c_wis_desc.GetStrides(), a_g_n_c_wis_strides);
copy(wei_g_k_c_xs_desc.GetLengths(), b_g_k_c_xs_lengths);
copy(wei_g_k_c_xs_desc.GetStrides(), b_g_k_c_xs_strides);
copy(out_g_n_k_wos_desc.GetLengths(), e_g_n_k_wos_lengths);
copy(out_g_n_k_wos_desc.GetStrides(), e_g_n_k_wos_strides);
copy(conv_param.conv_filter_strides_, conv_filter_strides);
copy(conv_param.conv_filter_dilations_, conv_filter_dilations);
copy(conv_param.input_left_pads_, input_left_pads);
copy(conv_param.input_right_pads_, input_right_pads);
// do Conv
auto conv = DeviceConvNDFwdInstance{};
auto invoker = conv.MakeInvoker();
auto argument = conv.MakeArgument(in_device_buf.GetDeviceBuffer(),
wei_device_buf.GetDeviceBuffer(),
std::array<const void*, 0>{},
out_device_buf.GetDeviceBuffer(),
a_g_n_c_wis_lengths,
a_g_n_c_wis_strides,
b_g_k_c_xs_lengths,
b_g_k_c_xs_strides,
std::array<std::array<ck::index_t, NDimSpatial + 3>, 0>{{}},
std::array<std::array<ck::index_t, NDimSpatial + 3>, 0>{{}},
e_g_n_k_wos_lengths,
e_g_n_k_wos_strides,
conv_filter_strides,
conv_filter_dilations,
input_left_pads,
input_right_pads,
in_element_op,
wei_element_op,
out_element_op);
if(!conv.IsSupportedArgument(argument))
{
throw std::runtime_error("The device op with the specified compilation parameters does "
"not support this convolution problem.");
}
float avg_time = invoker.Run(argument, StreamConfig{nullptr, time_kernel});
std::size_t flop = conv_param.GetFlops();
std::size_t num_btype = conv_param.GetByte<InDataType, WeiDataType, OutDataType>();
float tflops = static_cast<float>(flop) / 1.E9 / avg_time;
float gb_per_sec = num_btype / 1.E6 / avg_time;
std::cout << "Perf: " << avg_time << " ms, " << tflops << " TFlops, " << gb_per_sec << " GB/s, "
<< conv.GetTypeString() << std::endl;
if(do_verification)
{
auto ref_conv = ck::tensor_operation::host::ReferenceConvFwd<NDimSpatial,
InDataType,
WeiDataType,
OutDataType,
InElementOp,
WeiElementOp,
OutElementOp>();
auto ref_invoker = ref_conv.MakeInvoker();
auto ref_argument = ref_conv.MakeArgument(in,
wei,
out_host,
conv_param.conv_filter_strides_,
conv_param.conv_filter_dilations_,
conv_param.input_left_pads_,
conv_param.input_right_pads_,
in_element_op,
wei_element_op,
out_element_op);
ref_invoker.Run(ref_argument);
out_device_buf.FromDevice(out_device.mData.data());
return ck::utils::check_err(out_device, out_host, "Error: incorrect results!");
}
return true;
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_abs_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::UnaryAbs out_element_op;
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_clippedrelu_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::ClippedRelu out_element_op(0.f, 1.f);
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_elu_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::Elu out_element_op(2.f);
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_leakyrelu_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::LeakyRelu out_element_op(0.f);
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_logistic_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::Logistic out_element_op(1.0f);
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_passthrough_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::PassThrough out_element_op;
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_pow_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::Power out_element_op(4.f, 1.f, 2.f);
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_relu_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::Relu out_element_op;
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_sigmoid_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::Sigmoid out_element_op;
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_softrelu_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::SoftRelu out_element_op;
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_swish_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::Swish out_element_op(1.0f);
return !run_convnd_example(argc, argv, out_element_op);
}

13
example/62_convnd_activ/dynamic_unary/convnd_fwd_xdl_dynamic_tanh_fp16.cpp Normal file
View File

@@ -0,0 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#include "convnd_fwd_activ_dynamic_unary_common.hpp"
#include "../run_convnd_activ_dynamic_example.inc"
int main(int argc, char* argv[])
{
ck::tensor_operation::element_wise::TanH out_element_op;
return !run_convnd_example(argc, argv, out_element_op);
}

91
example/62_convnd_activ/run_convnd_activ_dynamic_example.inc Normal file
View File

@@ -0,0 +1,91 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2023-2024, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
void print_helper_msg()
{
std::cout << "arg1: verification (0=no, 1=yes)\n"
<< "arg2: initialization (0=no init, 1=integer value, 2=decimal value)\n"
<< "arg3: time kernel (0=no, 1=yes)\n"
<< ck::utils::conv::get_conv_param_parser_helper_msg() << std::endl;
}
template <typename OutElementOp>
bool run_convnd_example(int argc, char* argv[], const OutElementOp& out_element_op)
{
print_helper_msg();
bool do_verification = true;
// Use floats for SoftRelu by default to avoid overflow after e^x.
int init_method =
std::is_same_v<OutElementOp, ck::tensor_operation::element_wise::SoftRelu> ? 2 : 1;
bool time_kernel = false;
// Following shapes are selected to avoid overflow. Expect inf in case of
// size increase for some elementwise ops.
ck::utils::conv::ConvParam conv_param{
3, 2, 16, 128, 8, {3, 3, 3}, {17, 17, 17}, {2, 2, 2}, {1, 1, 1}, {1, 1, 1}, {1, 1, 1}};
if(argc == 1)
{
// use default
}
else if(argc == 4)
{
do_verification = std::stoi(argv[1]);
init_method = std::stoi(argv[2]);
time_kernel = std::stoi(argv[3]);
}
else
{
do_verification = std::stoi(argv[1]);
init_method = std::stoi(argv[2]);
time_kernel = std::stoi(argv[3]);
const ck::index_t num_dim_spatial = std::stoi(argv[4]);
conv_param = ck::utils::conv::parse_conv_param(num_dim_spatial, 5, argv);
}
const auto in_element_op = InElementOp{};
const auto wei_element_op = WeiElementOp{};
const auto run = [&]() {
const auto in_g_n_c_wis_desc =
ck::utils::conv::make_input_host_tensor_descriptor_g_n_c_wis_packed<InLayout>(
conv_param);
const auto wei_g_k_c_xs_desc =
ck::utils::conv::make_weight_host_tensor_descriptor_g_k_c_xs_packed<WeiLayout>(
conv_param);
const auto out_g_n_k_wos_desc =
ck::utils::conv::make_output_host_tensor_descriptor_g_n_k_wos_packed<OutLayout>(
conv_param);
return run_grouped_conv<NDimSpatial,
InDataType,
WeiDataType,
OutDataType,
InElementOp,
WeiElementOp,
OutElementOp,
DeviceGroupedConvNDActivInstance>(do_verification,
init_method,
time_kernel,
conv_param,
in_g_n_c_wis_desc,
wei_g_k_c_xs_desc,
out_g_n_k_wos_desc,
in_element_op,
wei_element_op,
out_element_op);
};
if(conv_param.num_dim_spatial_ == 3)
{
return run();
}
return false;
}