[CK_BUILDER] Add grouped conv fwd ck tile profiler (#3518)

* [BULDER] Add grouped conv fwd ck tile profiler

* [CK TILE] Fix grouped conv kernels splitk and double lds

* Updates

* Fixes

* Move to ckProfiler

* Fixes

* fix

* fix

* Change instances to empty list by default

* fix

* fix

* Update grouped_convolution_signatures.hpp

* Update grouped_convolution_forward_tile_algs.hpp

* [CK TILE] Add grouped convolution forward tests (#3556)

* [CK TILE] Add grouped convolution forward tests

* fix jenkins

* fixes

* comments fixes

* unit test

* unit test fix

* Move instances outside builder

* fix includes

* clang format fix

* readme fix

* fix includes

* fixes
This commit is contained in:
Bartłomiej Kocot
2026-01-20 06:29:01 +01:00
committed by GitHub
parent 0517d43d31
commit 0727e85e52
44 changed files with 3083 additions and 65 deletions

View File

@@ -98,27 +98,26 @@ struct ConvTileFactory
using GemmPipeline = typename internal::TilePipelineType<
BLOCK_GEMM.pipeline_version>::template GemmPipeline<UniversalGemmProblem>;
using ConvEpilogue = ck_tile::CShuffleEpilogue<ck_tile::CShuffleEpilogueProblem<
typename Types::ADataType,
typename Types::BDataType,
typename Types::DsDataTypes,
typename Types::AccDataType,
typename Types::EDataType,
typename GroupedConvTraitsType::ImplicitGemmDsLayout,
typename GroupedConvTraitsType::FixedGemmParams::ELayout,
typename Ops::CDEElementwiseOp,
BLOCK.per_block.m,
BLOCK.per_block.n,
BLOCK_GEMM.warps.m,
BLOCK_GEMM.warps.n,
BLOCK_GEMM.warp_tile.m,
BLOCK_GEMM.warp_tile.n,
BLOCK_GEMM.warp_tile.k,
GroupedConvTraitsType::FixedGemmParams::TransposeC,
// TODO:: This template parameter will be moved inside the kernel
BLOCK_GEMM.num_wave_groups,
GroupedConvTraitsType::FixedGemmParams::FixedVectorSize,
SCALAR_PER_VECTOR.c>>;
using ConvEpilogue = ck_tile::CShuffleEpilogue<
ck_tile::CShuffleEpilogueProblem<typename Types::ADataType,
typename Types::BDataType,
typename Types::DsDataTypes,
typename Types::AccDataType,
typename Types::EDataType,
typename GroupedConvTraitsType::ImplicitGemmDsLayout,
typename GroupedConvTraitsType::FixedGemmParams::ELayout,
typename Ops::CDEElementwiseOp,
BLOCK.per_block.m,
BLOCK.per_block.n,
BLOCK_GEMM.warps.m,
BLOCK_GEMM.warps.n,
BLOCK_GEMM.warp_tile.m,
BLOCK_GEMM.warp_tile.n,
BLOCK_GEMM.warp_tile.k,
GroupedConvTraitsType::FixedGemmParams::TransposeC,
BLOCK_GEMM.num_wave_groups,
GroupedConvTraitsType::FixedGemmParams::FixedVectorSize,
SCALAR_PER_VECTOR.c>>;
using Instance = typename internal::GroupedConvolutionTileKernel<SIGNATURE,
GroupedConvTraitsType,

View File

@@ -10,6 +10,7 @@
#include "ck_tile/builder/testing/testing_reflect.hpp"
#include "ck_tile/builder/testing/filter_extent.hpp"
#include "ck_tile/builder/testing/tensor_buffer.hpp"
#include "ck_tile/host/convolution_parameter.hpp"
#include "ck_tile/builder/testing/tensor_initialization.hpp"
#include "ck_tile/builder/testing/tensor_descriptor.hpp"
#include "ck_tile/builder/testing/validation.hpp"
@@ -93,6 +94,8 @@ struct Args<SIGNATURE>
Ops::WeiElementwiseOp b_elementwise_op;
Ops::OutElementwiseOp cde_elementwise_op;
int k_batch = 1;
/// This function returns the `TensorDescriptor` corresponding to
/// the input-tensor of the convolution problem. This can then
/// be used to, for example, allocate memory.
@@ -169,6 +172,36 @@ struct Args<SIGNATURE>
to_vector(this->input_left_pad),
to_vector(this->input_right_pad));
}
/// Convert the Args structure into a CK Tile conv_param structure.
/// This function is mainly used to be able to use the existing
/// CK Tile functionality to obtain tensor descriptors.
ck_tile::conv::ConvParam to_ck_tile_conv_param() const
{
const auto to_vector = [](const auto& extent) {
if constexpr(SPATIAL_DIM == 1)
return std::vector<ck_tile::index_t>{ck::index_t(extent.width)};
else if constexpr(SPATIAL_DIM == 2)
return std::vector<ck_tile::index_t>{ck::index_t(extent.height),
ck::index_t(extent.width)};
else
return std::vector<ck_tile::index_t>{ck::index_t(extent.depth),
ck::index_t(extent.height),
ck::index_t(extent.width)};
};
return ck_tile::conv::ConvParam(SPATIAL_DIM,
this->lengths.groups,
this->lengths.batch_size,
this->lengths.output_channels,
this->lengths.input_channels,
to_vector(this->lengths.filter),
to_vector(this->lengths.image),
to_vector(this->filter_strides),
to_vector(this->filter_dilation),
to_vector(this->input_left_pad),
to_vector(this->input_right_pad));
}
};
/// @brief `Inputs` specialization for forward convolution.

View File

@@ -4,6 +4,7 @@
#pragma once
#include "ck_tile/builder/testing/conv_fwd.hpp"
#include "ck_tile/host/kernel_launch.hpp"
#include "ck_tile/builder/factory/helpers/ck/conv_elementwise_op.hpp"
#include <type_traits>
#include <array>
@@ -87,16 +88,19 @@ concept CkConvInstance = detail::CkConvInstance<Conv, SIGNATURE>;
/// @brief `run()` specialization for forward convolution and old CK.
///
/// @tparam SIGNATURE Forward convolution signature.
/// @throws std::runtime_error if the arguments werent actually valid for the
/// @throws std::runtime_error if the arguments weren't actually valid for the
/// operation. This should be caught and reported by the testing framework.
/// @return std::tuple<bool, float> - whether the problem is supported and
/// kernel execution time (0.0f if s_conf time_kernel is false).
///
/// @see run()
template <auto SIGNATURE>
requires ValidConvSignature<SIGNATURE> && ConvDirectionIsForward<SIGNATURE>
void run(CkConvInstance<SIGNATURE> auto& conv,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs)
std::tuple<bool, float> run(CkConvInstance<SIGNATURE> auto& conv,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs,
const StreamConfig s_conf = {})
{
constexpr auto spatial_dim = SIGNATURE.spatial_dim;
@@ -144,10 +148,10 @@ void run(CkConvInstance<SIGNATURE> auto& conv,
if(!conv.IsSupportedArgument(ck_args))
{
throw std::runtime_error("invalid argument");
std::cout << "invalid argument" << std::endl;
}
conv.MakeInvoker().Run(ck_args, {});
return std::make_tuple(true, conv.MakeInvoker().Run(ck_args, s_conf));
}
} // namespace ck_tile::builder::test

View File

@@ -0,0 +1,91 @@
// Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
// SPDX-License-Identifier: MIT
#pragma once
#include "ck_tile/builder/testing/conv_fwd.hpp"
#include "ck_tile/host/kernel_launch.hpp"
#include "ck_tile/builder/factory/helpers/ck/conv_elementwise_op.hpp"
#include "ck_tile/ops/gemm.hpp"
#include "ck_tile/ops/grouped_convolution.hpp"
#include <type_traits>
#include <array>
/// This file contains the implementation details for invoking/testing
/// grouped convolution operations in CK Tile. The main item is the
/// `run()` function, which is the main implementation used to invoke
/// CK Tile grouped forward convolution kernels.
namespace ck_tile::builder::test {
namespace detail {
/// @brief Concept for checking whether this is the CK Tile convolution
/// implementation.
///
/// This is the same as `::ck_tile::builder::test::CkConvInstance`, except
/// with some utility aliases. For that reason, its moved to this detail
/// namespace.
template <typename Conv, auto SIGNATURE>
concept CkTileConvInstance = requires(Conv&) {
{ Conv::BlockSize() };
};
} // namespace detail
/// @brief Concept for checking whether a convolution is invoked like CK Tile.
///
/// This concept is used to tell whether a convolution implementation is
/// likely to be an "CK Tile" implementation - that is, whether we should
/// invoke it as an CK Tile kernel. This is mainly used with `run()` to
/// differentiate which implementation that should be invoked.
///
/// - SIGNATURE is the operation signature.
/// - Conv is a convolution instance created by the CK Builder API.
template <typename Conv, auto SIGNATURE>
concept CkTileConvInstance = detail::CkTileConvInstance<Conv, SIGNATURE>;
/// @brief `run()` specialization for forward convolution and CK Tile.
///
/// @tparam SIGNATURE Forward convolution signature.
/// @throws std::runtime_error if the arguments weren't actually valid for the
/// operation. This should be caught and reported by the testing framework.
/// @return std::tuple<bool, float> - whether the problem is supported and
/// kernel execution time (0.0f if s_conf time_kernel is false).
///
/// @see run()
template <auto SIGNATURE>
requires ValidConvSignature<SIGNATURE> && ConvDirectionIsForward<SIGNATURE>
std::tuple<bool, float> run(CkTileConvInstance<SIGNATURE> auto& conv,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs,
const ck_tile::stream_config s_conf = {})
{
using Conv = std::remove_reference_t<decltype(conv)>;
const auto param = args.to_ck_tile_conv_param();
ck_tile::GroupedConvFwdHostArgs<> host_args(
param, inputs.input, inputs.weight, {}, outputs.output, args.k_batch);
auto kargs = Conv::MakeKernelArgs(host_args);
const dim3 grids = Conv::GridSize(kargs);
const dim3 blocks = Conv::BlockSize();
if(!Conv::IsSupportedArgument(kargs))
{
std::cout << "Not supported!";
return std::make_tuple(false, 0.f);
}
constexpr index_t minimum_occupancy =
Conv::GemmPipeline::Scheduler == ck_tile::GemmPipelineScheduler::Intrawave ? 1 : 2;
return std::make_tuple(
true,
ck_tile::launch_kernel(
s_conf, ck_tile::make_kernel<minimum_occupancy>(conv, grids, blocks, 0, kargs)));
}
} // namespace ck_tile::builder::test

View File

@@ -62,6 +62,8 @@ concept RefConvInstance = requires(Conv& conv,
/// @throws std::runtime_error if the arguments weren't actually valid for the
/// operation. This should be caught and reported by the testing framework.
///
/// @return std::tuple<bool, float> - whether the problem is supported and
/// kernel execution time (0.0f for reference).
/// @see run()
template <auto SIGNATURE>
requires ValidConvSignature<SIGNATURE> &&
@@ -69,10 +71,10 @@ template <auto SIGNATURE>
// for now, just concern outselves with reference and see when the
// rest of the bwd/weight plumbing is there.
ConvDirectionIsForward<SIGNATURE>
void run(RefConvInstance<SIGNATURE> auto& conv,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs)
std::tuple<bool, float> run(RefConvInstance<SIGNATURE> auto& conv,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs)
{
// We don't want to compute the output dims manually, just get
// them via the existing infrastructure
@@ -86,15 +88,27 @@ void run(RefConvInstance<SIGNATURE> auto& conv,
for(auto right_pad : param.input_right_pads_)
{
if(right_pad != 0)
throw std::runtime_error("TODO: Support right pad in reference conv");
{
std::cout << "TODO: Support right pad in reference conv" << std::endl;
return std::make_tuple(false, 0.0f);
}
}
if(!args.make_input_descriptor().is_packed())
throw std::runtime_error("TODO: Support non-packed input tensor in reference conv");
{
std::cout << "TODO: Support non-packed input tensor in reference conv" << std::endl;
return std::make_tuple(false, 0.0f);
}
if(!args.make_weight_descriptor().is_packed())
throw std::runtime_error("TODO: Support non-packed weight tensor in reference conv");
{
std::cout << "TODO: Support non-packed weight tensor in reference conv" << std::endl;
return std::make_tuple(false, 0.0f);
}
if(!args.make_output_descriptor().is_packed())
throw std::runtime_error("TODO: Support non-packed output tensor in reference conv");
{
std::cout << "TODO: Support non-packed output tensor in reference conv" << std::endl;
return std::make_tuple(false, 0.0f);
}
conv.Run(inputs.input,
inputs.weight,
@@ -109,6 +123,7 @@ void run(RefConvInstance<SIGNATURE> auto& conv,
param.conv_filter_strides_,
param.conv_filter_dilations_,
param.input_left_pads_);
return std::make_tuple(true, 0.0f);
}
} // namespace ck_tile::builder::test

View File

@@ -34,4 +34,25 @@ struct FilterExtent<3>
size_t depth = 1;
};
template <int SPATIAL_DIM>
inline FilterExtent<SPATIAL_DIM> filter_extent_from_vector(const std::vector<std::size_t>& vec);
template <>
inline FilterExtent<1> filter_extent_from_vector<1>(const std::vector<std::size_t>& vec)
{
return FilterExtent<1>{.width = vec[0]};
}
template <>
inline FilterExtent<2> filter_extent_from_vector<2>(const std::vector<std::size_t>& vec)
{
return FilterExtent<2>{.width = vec[1], .height = vec[0]};
}
template <>
inline FilterExtent<3> filter_extent_from_vector<3>(const std::vector<std::size_t>& vec)
{
return FilterExtent<3>{.width = vec[2], .height = vec[1], .depth = vec[0]};
}
} // namespace ck_tile::builder::test

View File

@@ -418,6 +418,10 @@ struct TensorDescriptor
size_t x = 1;
for(size_t i = 0; i < RANK; ++i)
{
if(lengths[indices[i]] == 1)
{
continue;
}
if(strides[indices[i]] != x)
return false;
@@ -443,6 +447,15 @@ struct TensorDescriptor
return TensorDescriptor<DT, 1>(lengths, strides);
}
/// @brief Print tensor descriptor details.
///
/// Print tensor descriptor details - lengths and strides.
friend std::ostream& operator<<(std::ostream& os, const TensorDescriptor<DT, RANK>& tensor_desc)
{
os << tensor_desc.inner_descriptor_;
return os;
}
private:
ck_tile::HostTensorDescriptor inner_descriptor_;
};

View File

@@ -317,13 +317,17 @@ ValidationReport validate(const Args<SIGNATURE>& args,
/// @param inputs The input tensor data. Will not be modified by this function.
/// @param outputs The output tensor data. The contents will be overwritten by
/// this function.
/// @param s_conf Stream config used to launch kernel.
/// @return std::tuple<bool, float> - whether the problem is supported and
/// kernel execution time (0.0f if s_conf time_kernel is false).
///
/// @note This function is explicitly deleted to generate compile errors
/// for missing implementations.
template <auto SIGNATURE, typename Operation>
void run(Operation& operation,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs) = delete;
template <auto SIGNATURE, typename Operation, typename StreamConf>
std::tuple<bool, float> run(Operation& operation,
const Args<SIGNATURE>& args,
const Inputs<SIGNATURE>& inputs,
const Outputs<SIGNATURE>& outputs,
const StreamConf s_conf = {}) = delete;
} // namespace ck_tile::builder::test

View File

@@ -0,0 +1,84 @@
// Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
// SPDX-License-Identifier: MIT
#include "utils/ckb_conv_tile_test_configs.hpp"
#include "utils/ckb_conv_test_utils.hpp"
#include "utils/conv_algorithm_type_utils.hpp"
#include "ck_tile/builder/testing/conv_fwd_ck_tile.hpp"
#include "ck_tile/builder/testing/conv_fwd_reference.hpp"
#include "ck_tile/host/device_prop.hpp"
#include "testing_utils.hpp"
namespace ckb = ck_tile::builder;
namespace ckt = ck_tile::builder::test;
namespace cku = ck_tile::builder::test_utils;
constexpr auto SIGNATURE =
ckt::ConvSignature{.spatial_dim = 2,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::FP16,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NHWGK}}};
constexpr auto ALGORITHM =
cku::ConvAlgorithm_Tile_GroupedConvolutionKernel{}
.with_tile_specializations(ckb::TileConvSpecialization::DEFAULT)
.with_tile_thread_block(cku::FwdTileThreadBlock_64x64x64)
.with_tile_block_gemm(cku::TileBlockGemmDesc_16x16_v3_intrawave)
.with_tile_transfer(cku::FwdTileTransfer_4x4x4)
.with_tile_optimizations(ckt::TileOptimizations{
.num_groups_to_merge = 1, .split_image = false, .explicit_gemm = false});
using Builder = ckb::ConvBuilder<SIGNATURE, ALGORITHM>;
using Instance = Builder::Instance;
using Reference = ckb::ConvBuilder<SIGNATURE, ckt::ConvAlgorithm_Reference{}>::Instance;
TEST(Fwd2DFp16_CShufV3_NHWGC, EndToEnd)
{
if(!ck_tile::get_device_name().starts_with("gfx9"))
{
GTEST_SKIP() << "unsupported architecture";
}
ckt::Args<SIGNATURE> args = {
.lengths =
{
.batch_size = 16,
.groups = 1,
.input_channels = 32,
.output_channels = 48,
.image =
{
.width = 56,
.height = 64,
},
.filter =
{
.width = 3,
.height = 5,
},
},
.filter_strides = {.width = 1, .height = 1},
.filter_dilation = {.width = 1, .height = 1},
.input_left_pad = {.width = 0, .height = 0},
.input_right_pad = {.width = 0, .height = 0},
.a_elementwise_op = {},
.b_elementwise_op = {},
.cde_elementwise_op = {},
};
auto inputs = alloc_inputs(args);
auto outputs = alloc_outputs(args);
auto reference = alloc_outputs(args);
ckt::init_inputs(args, inputs.get());
auto conv = Instance{};
ckt::run(conv, args, inputs.get(), outputs.get());
auto ref_conv = Reference{};
ckt::run(ref_conv, args, inputs.get(), reference.get());
EXPECT_THAT(outputs.get(), ck_tile::test::MatchesReference(args, reference.get()));
}

View File

@@ -16,6 +16,8 @@ struct TensorConfig
// Optional data types, override the type defined in the signature if provided.
DataType data_type{DataType::UNDEFINED_DATA_TYPE};
DataType compute_type{DataType::UNDEFINED_DATA_TYPE};
constexpr bool operator==(const TensorConfig& other) const = default;
};
template <TensorConfig... Configs>
@@ -31,6 +33,8 @@ struct TensorOperation
return TensorOperation<Configs..., TensorConfig{AuxiliaryConfigs}...>{
.elementwise_operation = this->elementwise_operation};
}
constexpr bool operator==(const TensorOperation& other) const = default;
};
template <typename Op = TensorOperation<>>
@@ -38,6 +42,8 @@ struct ConvolutionTensor
{
TensorConfig config;
Op operation{};
constexpr bool operator==(const ConvolutionTensor& other) const = default;
};
template <typename InputTensor = ConvolutionTensor<>,
@@ -52,6 +58,8 @@ struct ConvSignature
InputTensor input;
WeightTensor weight;
OutputTensor output;
constexpr bool operator==(const ConvSignature& other) const = default;
};
} // namespace ck_tile::builder::test

View File

@@ -190,6 +190,8 @@ TEST(TensorDescriptor, IsPacked)
ckt::make_descriptor<dt>(ckt::Extent{10, 11, 12}, ckt::Extent{1, 100, 1100}).is_packed());
EXPECT_FALSE(
ckt::make_descriptor<dt>(ckt::Extent{30, 20, 10}, ckt::Extent{1, 1, 1}).is_packed());
EXPECT_TRUE(
ckt::make_descriptor<dt>(ckt::Extent{30, 20, 1}, ckt::Extent{1, 30, 30}).is_packed());
}
TEST(TensorDescriptor, PrintExtent)

View File

@@ -3,8 +3,8 @@
#pragma once
#include "impl/conv_algorithm_types.hpp"
#include "impl/conv_signature_types.hpp"
#include "../impl/conv_algorithm_types.hpp"
#include "../impl/conv_signature_types.hpp"
#include "ck_tile/builder/conv_builder.hpp"
namespace ck_tile::builder::test_utils {

View File

@@ -0,0 +1,19 @@
# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
# SPDX-License-Identifier: MIT
if(GPU_TARGETS MATCHES "gfx9")
# Generate instances using python script (empty to just generate empty instance list)
if(NOT EXISTS ${CMAKE_CURRENT_SOURCE_DIR}/grouped_convolution_forward_tile_ndhwgc_fp32.inc)
find_package(Python3 COMPONENTS Interpreter Development)
execute_process(
COMMAND ${Python3_EXECUTABLE} ${CMAKE_CURRENT_SOURCE_DIR}/generate_instances.py --mode=tests
RESULT_VARIABLE ret
)
endif()
# Find cpp files and create lib for instances
file(GLOB_RECURSE GROUPED_CONV_FWD_TILE "instances/*.cpp")
add_instance_library(device_grouped_conv_fwd_tile_instances ${GROUPED_CONV_FWD_TILE})
target_include_directories(device_grouped_conv_fwd_tile_instances PRIVATE
"${PROJECT_SOURCE_DIR}/experimental/builder/test/utils")
endif()

View File

@@ -0,0 +1,5 @@
# Grouped Convolution Tile Instances Generator
CK Tile Convolution instances implemented via builder and generated via python script.
It is integrated with tests and ckProfiler
This functionality will be refactored and moved under the Tile Engine.
At now to speed up development and provide tests for CK Tile Convolution it has been implemented under experimental directory.

View File

@@ -0,0 +1,237 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 2, 2, 2, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Default, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Stride1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Default, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Stride1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 64, 64, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 64, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 64, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,228 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 2, 2, 2, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Default, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Stride1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Default, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Stride1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,176 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Default, 32, 32, 2, 1, 4, 4, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Default, 32, 32, 2, 2, 1, 1, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 16, Default, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 16, Default, 32, 32, 2, 4, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 16, Default, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 16, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 16, Default, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 16, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 16, Default, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 16, Default, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 192, 16, Default, 32, 32, 2, 3, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 16, Filter1x1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 16, Filter1x1Pad0, 32, 32, 2, 4, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 16, Filter1x1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 16, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 16, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 16, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 192, 16, Filter1x1Pad0, 32, 32, 2, 3, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 16, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 16, Filter1x1Stride1Pad0, 32, 32, 2, 4, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 16, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 16, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 16, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 16, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 192, 16, Filter1x1Stride1Pad0, 32, 32, 2, 3, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 16, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 16, Default, 32, 32, 4, 2, 4, 4, 4, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,237 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 2, 2, 2, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 64, 64, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 64, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 64, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 64, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Default, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Default, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Stride1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Stride1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,228 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Default, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 32, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 2, 2, 2, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Default, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Default, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Default, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
# DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 32, 32, 4, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 224, 256, 64, Filter1x1Stride1Pad0, 16, 16, 7, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 224, 64, Filter1x1Stride1Pad0, 16, 16, 8, 7, 8, 8, 8, 2, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 128, 32, Filter1x1Stride1Pad0, 32, 32, 4, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Default, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 16, 256, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 32, 256, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,176 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Default, 32, 32, 2, 1, 4, 4, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Default, 32, 32, 2, 2, 1, 1, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 16, Default, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 16, Default, 32, 32, 2, 4, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 16, Default, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Default, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 16, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 16, Default, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 16, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 16, Default, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 16, Default, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 192, 16, Default, 32, 32, 2, 3, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 16, Filter1x1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 16, Filter1x1Pad0, 32, 32, 2, 4, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 16, Filter1x1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 16, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 16, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 16, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 192, 16, Filter1x1Pad0, 32, 32, 2, 3, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 256, 128, 16, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 256, 16, Filter1x1Stride1Pad0, 32, 32, 2, 4, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 128, 16, Filter1x1Stride1Pad0, 32, 32, 4, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 128, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 64, 128, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 64, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 128, 16, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 128, 32, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<128, 32, 128, 16, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 32, 16, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 16, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 128, 192, 16, Filter1x1Stride1Pad0, 32, 32, 2, 3, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 2, 1, 2, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 4, 4, 4, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 16, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<256, 256, 128, 16, Default, 32, 32, 4, 2, 4, 4, 4, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Default, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Default, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Default, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Default, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Default, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Default, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Default, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Default, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 128, 16, 64, Filter1x1Stride1Pad0, 16, 16, 4, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 32, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 64, 16, 64, Filter1x1Stride1Pad0, 16, 16, 2, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 64, 64, Filter1x1Stride1Pad0, 16, 16, 1, 2, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 64, 64, Filter1x1Stride1Pad0, 32, 32, 1, 1, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 128, 64, Filter1x1Stride1Pad0, 16, 16, 1, 4, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 64, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,41 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,41 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,42 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 32, Default, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,41 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,41 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Default, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 32, 64, Filter1x1Stride1Pad0, 32, 32, 2, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,42 @@
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Default, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 64, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 1, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 32, 64, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 8, 8, 8, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Default, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<256, 64, 64, 32, Filter1x1Stride1Pad0, 16, 16, 2, 2, 1, 2, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleD_Xdl_CShuffle_Large_Tensor<64, 64, 64, 32, Default, 32, 32, 2, 2, 1, 1, 1, 1, 1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Default, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 8>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 16>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle<64, 64, 16, 16, Filter3x3, 16, 16, 4, 1, 4, 1, 1, 1, 1, 32>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 32, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v4>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Default, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 256, 256, 32, Filter1x1Stride1Pad0, 16, 16, 8, 8, 8, 8, 8, 1, 2, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v3>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Default, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 128, 64, Filter1x1Stride1Pad0, 32, 32, 2, 2, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v5>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Default, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<256, 128, 256, 32, Filter1x1Stride1Pad0, 32, 32, 2, 4, 8, 8, 8, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Default, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 16, 32, 64, Default, 16, 16, 1, 1, 8, 8, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 16, 64, Filter1x1Stride1Pad0, 16, 16, 1, 1, 8, 8, 2, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v1>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Default, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<64, 16, 16, 128, Filter1x1Stride1Pad0, 16, 16, 1, 1, 4, 4, 4, 1, 1, BlkGemmPipelineScheduler: Interwave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 32, Default, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 32, Filter1x1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>
DeviceGroupedConvFwdMultipleABD_Xdl_CShuffle_V3<128, 32, 128, 32, Filter1x1Stride1Pad0, 32, 32, 1, 2, 4, 4, 8, 1, 1, BlkGemmPipelineScheduler: Intrawave, BlkGemmPipelineVersion: v2>

View File

@@ -0,0 +1,275 @@
# Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
# SPDX-License-Identifier: MIT
import argparse
from pathlib import Path
class ConvInstanceTemplateParams:
def __init__(
self,
specialization,
tile_size,
warps,
warp_tile,
double_smem_buffer,
num_wave_groups,
pipeline_version,
scheduler,
scalar_per_vector,
num_groups_to_merge,
split_image,
explicit_gemm,
id,
):
self.specialization = specialization
self.tile_size = tile_size
self.warps = warps
self.warp_tile = warp_tile
self.double_smem_buffer = double_smem_buffer
self.num_wave_groups = num_wave_groups
self.pipeline_version = pipeline_version
self.scheduler = scheduler
self.scalar_per_vector = scalar_per_vector
self.num_groups_to_merge = num_groups_to_merge
self.split_image = split_image
self.explicit_gemm = explicit_gemm
self.id = id
def get_optimizations(self):
explicit_gemm = "true" if self.explicit_gemm else "false"
split_image = "true" if self.split_image else "false"
num_groups_to_merge = str(self.num_groups_to_merge)
return f"ckt::TileOptimizations{{.num_groups_to_merge = {num_groups_to_merge}, .split_image = {split_image}, .explicit_gemm = {explicit_gemm}}}"
def get_specialization(self):
namespace = "ckb::TileConvSpecialization::"
if self.specialization == "Default" or self.specialization == "OddC":
return namespace + "DEFAULT"
if self.specialization == "Filter1x1Pad0":
return namespace + "FILTER_1X1_PAD0"
if self.specialization == "Filter1x1Stride1Pad0":
return namespace + "FILTER_1X1_STRIDE1_PAD0"
if self.specialization == "Filter3x3":
return namespace + "FILTER_3x3"
else:
raise RuntimeError("not supported specialization")
def get_thread_block(self):
return f"ckt::TileThreadBlock{{.tile_size = {{.m = {self.tile_size[0]}, .n = {self.tile_size[1]}, .k = {self.tile_size[2]}}}}}"
def get_block_gemm_desc(self):
double_smem_buffer = "true" if self.double_smem_buffer else "false"
pipeline_version = self.pipeline_version[-1:]
scheduler = (
"INTRAWAVE" if self.scheduler.find("Intrawave") != -1 else "INTERWAVE"
)
return f"""ckt::TileBlockGemm{{
.warps = {{.m = {self.warps[0]}, .n = {self.warps[1]}, .k = {self.warps[2]}}},
.warp_tile = {{.m = {self.warp_tile[0]}, .n = {self.warp_tile[1]}, .k = {self.warp_tile[2]}}},
.double_smem_buffer = {double_smem_buffer},
.num_wave_groups = {self.num_wave_groups},
.pipeline_version = ckb::PipelineVersion::V{pipeline_version},
.scheduler = ckb::PipelineScheduler::{scheduler}}}"""
def get_block_transfer(self):
return f"""ckt::TileTransfer{{.a_scalar_per_vector = {self.scalar_per_vector[0]},
.b_scalar_per_vector = {self.scalar_per_vector[1]}, .c_scalar_per_vector = {self.scalar_per_vector[2]}}}"""
def get_dtype(problem_name):
if problem_name.find("fp32") != -1:
return "float"
if problem_name.find("fp16") != -1:
return "ck_tile::half_t"
if problem_name.find("bf16") != -1:
return "ck_tile::bf16_t"
else:
raise RuntimeError("wrong dtype")
def generate_calls_inc(instances, problem_name, direction, filter_pattern):
generate_dir = Path(__file__).resolve().parent
with open(f"{generate_dir}/{problem_name}_calls.inc", "w") as f:
if problem_name.find(filter_pattern) == -1:
return
for instance in instances:
instance_name = problem_name + "_" + str(instance.id)
f.write(f"run_alg(run_{instance_name});\n")
def generate_defs_inc(instances, problem_name, signature, direction, filter_pattern):
generate_dir = Path(__file__).resolve().parent
with open(f"{generate_dir}/{problem_name}.inc", "w") as f:
if problem_name.find(filter_pattern) == -1:
return
for instance in instances:
instance_name = problem_name + "_" + str(instance.id)
f.write(
f"std::tuple<bool, float, std::string> run_{instance_name}(\n"
f" const ckt::Args<{signature}>& args,\n"
f" const ckt::Inputs<{signature}>& inputs,\n"
f" const ckt::Outputs<{signature}>& outputs,\n"
f" const ck_tile::stream_config& s_conf);\n"
)
def generate_fwd_cpp(
instances, problem_name, config, direction, signature_name, filter_pattern
):
for instance in instances:
if problem_name.find(filter_pattern) == -1:
break
instance_name = problem_name + "_" + str(instance.id)
generate_dir = Path(__file__).resolve().parent
directory_path = Path(f"{generate_dir}/instances/{config}")
directory_path.mkdir(parents=True, exist_ok=True)
with open(
f"{generate_dir}/instances/grouped_convolution_forward_tile.cpp.in",
"r",
) as f:
content = f.read()
content = content.replace("gen_signature", signature_name)
content = content.replace("gen_instance_name", instance_name)
content = content.replace("gen_specialization", instance.get_specialization())
content = content.replace("gen_thread_block", instance.get_thread_block())
content = content.replace("gen_block_gemm_desc", instance.get_block_gemm_desc())
content = content.replace("gen_block_transfer", instance.get_block_transfer())
content = content.replace("gen_optimizations", instance.get_optimizations())
with open(
f"{generate_dir}/instances/{config}/{instance_name}.cpp",
"w",
) as f:
f.write(content)
def parse_fwd_instances(instances, problem_name):
convs = []
for instance_id, instance in enumerate(instances):
if instance.find("#") != -1 or instance.find(";") != -1:
continue
instance_args_list = instance[instance.find("<") + 1 : instance.find(">")]
args = instance_args_list.split(", ")
block_size = int(args[0])
m_per_block = int(args[1])
n_per_block = int(args[2])
k_per_block = int(args[3])
spec = args[4]
m_per_xdl = int(args[5])
n_per_xdl = int(args[6])
m_xdl_per_wave = int(args[7])
n_xdl_per_wave = int(args[8])
a_scalar_per_vector = int(args[9])
b_scalar_per_vector = int(args[10])
c_scalar_per_vector = int(args[11])
if len(args) == 15:
num_groups_to_merge = int(args[14])
elif len(args) != 16 and len(args) != 14:
raise RuntimeError("wrong number of parameters")
else:
num_groups_to_merge = 1
split_image = instance.find("Large") != -1
double_smem_buffer = instance.find("BlkGemmPipelineVersion: v4") != -1
num_wave_groups = 2 if instance.find("BlkGemmPipelineVersion: v5") != -1 else 1
scheduler = (
"Intrawave" if instance.find("BlkGemmPipelineScheduler") == -1 else args[14]
)
pipeline_version = (
"v1" if instance.find("BlkGemmPipelineVersion") == -1 else args[15]
)
m_warp = int(m_per_block / (m_per_xdl * m_xdl_per_wave))
n_warp = int(n_per_block / (n_per_xdl * n_xdl_per_wave))
warp_size = 64
k_warp = int(block_size / (warp_size * m_warp * n_warp))
dtype = get_dtype(problem_name)
# TODO: Make it more flexible
# k_per_xdl = f"ck_tile::get_k_warp_tile<{dtype}, {m_per_xdl}>()"
k_per_xdl = 8 if dtype == "float" else 16
conv = ConvInstanceTemplateParams(
spec,
[m_per_block, n_per_block, k_per_block],
[m_warp, n_warp, k_warp],
[m_per_xdl, n_per_xdl, k_per_xdl],
double_smem_buffer,
num_wave_groups,
pipeline_version,
scheduler,
[a_scalar_per_vector, b_scalar_per_vector, c_scalar_per_vector],
num_groups_to_merge,
split_image,
False,
instance_id,
)
convs.append(conv)
return convs
def generate_instances_fwd(instances, problem_name, config, filter_pattern):
direction = "forward"
signature_name = f"SIGNATURE_{config.upper()}_FWD"
instances = parse_fwd_instances(instances, problem_name)
generate_calls_inc(instances, problem_name, direction, filter_pattern)
generate_defs_inc(
instances,
problem_name,
signature_name,
direction,
filter_pattern,
)
generate_fwd_cpp(
instances, problem_name, config, direction, signature_name, filter_pattern
)
if __name__ == "__main__":
fwd_configs = [
"nhwgc_fp32",
"nhwgc_fp16",
"nhwgc_bf16",
"ndhwgc_fp32",
"ndhwgc_fp16",
"ndhwgc_bf16",
]
parser = argparse.ArgumentParser(
description="Generate grouped conv CK Tile instances."
)
parser.add_argument(
"--filter_pattern",
type=str,
default="convolution",
help="Filter pattern for configs.",
)
parser.add_argument(
"--mode",
choices=["compilation", "tests", "profiler"],
type=str,
default="profiler",
help="Generator modes. compilation - empty instance list, tests - limited instance list, profiler - generate all instances",
)
args = parser.parse_args()
# apply empty filter
if args.mode == "compilation":
args.filter_pattern = "empty"
configs_prefix = "profiler"
elif args.mode == "tests":
configs_prefix = "tests"
elif args.mode == "profiler":
configs_prefix = "profiler"
else:
raise RuntimeError("wrong mode")
for config in fwd_configs:
instances = []
generate_dir = Path(__file__).resolve().parent
config_path = f"{generate_dir}/configs/{configs_prefix}/{config}.conf"
with open(config_path, "r") as file:
instances = file.readlines()
problem_name = f"grouped_convolution_forward_tile_{config}"
generate_instances_fwd(instances, problem_name, config, args.filter_pattern)

View File

@@ -0,0 +1,19 @@
// Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
// SPDX-License-Identifier: MIT
#include "../instance_includes.inc"
namespace ck_tile::builder::profiling {
constexpr auto SIGNATURE = gen_signature;
std::tuple<bool, float, std::string> run_gen_instance_name(const ckt::Args<SIGNATURE>& args,
const ckt::Inputs<SIGNATURE>& inputs,
const ckt::Outputs<SIGNATURE>& outputs,
const ck_tile::stream_config& s_conf)
{
constexpr auto ALGORITHM = cku::ConvAlgorithm_Tile_GroupedConvolutionKernel{}
.with_tile_specializations(gen_specialization)
.with_tile_thread_block(gen_thread_block)
.with_tile_block_gemm(gen_block_gemm_desc)
.with_tile_transfer(gen_block_transfer)
.with_tile_optimizations(gen_optimizations);
#include "../instance_run.inc"
}
} // namespace ck_tile::builder::profiling

View File

@@ -0,0 +1,64 @@
#include "../../builder/test/utils/ckb_conv_tile_test_configs.hpp"
#include "ck_tile/builder/testing/conv_fwd_ck_tile.hpp"
namespace ckb = ck_tile::builder;
namespace ckt = ck_tile::builder::test;
namespace cku = ck_tile::builder::test_utils;
namespace ck_tile::builder::profiling {
constexpr auto SIGNATURE_NHWGC_FP32_FWD =
ckt::ConvSignature{.spatial_dim = 2,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::FP32,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NHWGK}}};
constexpr auto SIGNATURE_NHWGC_BF16_FWD =
ckt::ConvSignature{.spatial_dim = 2,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::BF16,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NHWGK}}};
constexpr auto SIGNATURE_NHWGC_FP16_FWD =
ckt::ConvSignature{.spatial_dim = 2,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::FP16,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NHWGK}}};
constexpr auto SIGNATURE_NDHWGC_FP32_FWD =
ckt::ConvSignature{.spatial_dim = 3,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::FP32,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NDHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKZYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NDHWGK}}};
constexpr auto SIGNATURE_NDHWGC_BF16_FWD =
ckt::ConvSignature{.spatial_dim = 3,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::BF16,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NDHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKZYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NDHWGK}}};
constexpr auto SIGNATURE_NDHWGC_FP16_FWD =
ckt::ConvSignature{.spatial_dim = 3,
.direction = ckb::ConvDirection::FORWARD,
.data_type = ckb::DataType::FP16,
.accumulation_data_type = ckb::DataType::FP32,
.input = {.config = {.layout = ckb::TensorLayout::NDHWGC}},
.weight = {.config = {.layout = ckb::TensorLayout::GKZYXC}},
.output = {.config = {.layout = ckb::TensorLayout::NDHWGK}}};
} // namespace ck_tile::builder::profiling

View File

@@ -0,0 +1,9 @@
using Builder = ckb::ConvBuilder<SIGNATURE, ALGORITHM>;
using Instance = Builder::Instance;
auto conv = Instance{};
bool is_supported;
float avg_time;
std::tie(is_supported, avg_time) = ckt::run(conv, args, inputs, outputs, s_conf);
return std::make_tuple(is_supported, avg_time, conv.GetInstanceString());