ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-17 11:30:02 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
ce838e19e5ea1931041537112d85185bb7ce7fa7
composable_kernel/profiler/src/profile_grouped_conv_bwd_weight.cpp
Ville Pietilä 1e8cde4b6d [CK_TILE] Add CK Tile bwd weight profiler (#4797) 
## Motivation

To compare old CK and CK Tile, we need to extend the current CK profiler
to support running also CK Tile instance with the same API. In order to
have the same instance coverage in CK Tile compared to the old CK, I've
added code generation from old CK configurations to CK Tile instances
using the CK Builder.

## Technical Details

- The codegen python script for CK Tile fwd convs is extended to support
also bwd weight and bwd data.
- The generated instances are added to the CMake build (target
`device_grouped_conv_bwd_weight_tile_instance`s).
- A new profiler op (`grouped_conv_bwd_weight_tile`) has been added to
the CK Profiler.

---------

Co-authored-by: Ville Pietilä <>
Co-authored-by: Bartlomiej Kocot <barkocot@amd.com>
2026-03-04 21:49:42 +00:00

647 lines
26 KiB
C++
Raw Blame History

// Copyright (c) Advanced Micro Devices, Inc., or its affiliates.
// SPDX-License-Identifier: MIT
#include <cstdlib>
#include <initializer_list>
#include <iostream>
#include <numeric>
#include "profiler/profile_grouped_conv_bwd_weight_impl.hpp"
#include "profiler/profiler_arg_utils.hpp"
#include "profiler_operation_registry.hpp"
namespace {
enum struct ConvLayout
{
GNCHW_GKCYX_GNKHW, // 0
GNHWC_GKYXC_GNHWK, // 1
NHWGC_GKYXC_NHWGK, // 2
NGCHW_GKYXC_NGKHW, // 3
NGCHW_GKCYX_NGKHW, // 4
};
enum struct ConvDataType
{
F32_F32_F32, // 0
F16_F16_F16, // 1
BF16_F32_BF16, // 2
F16_F16_F16_BF8_F8, // 3
I8_I8_I8, // 4
BF16_BF16_BF16, // 5
F32_F32_F32_TF32, // 6
};
#define OP_NAME "grouped_conv_bwd_weight"
#define OP_DESC "Grouped Convolution Backward Weight"
static void print_helper_msg()
{
std::cout << "arg1: tensor operation (" OP_NAME ": " OP_DESC ")\n"
<< "arg2: data type (0: Input fp32, Weight fp32, Output fp32\n"
<< " 1: Input fp16, Weight fp16, Output fp16\n"
<< " 2: Input bf16, Weight fp32, Output bf16\n"
<< " 3: Input fp16, Weight fp16, Output fp16, Gemm bf8@fp8\n"
<< " 4: Input int8, Weight int8, Output int8\n"
<< " 5: Input bf16, Weight bf16, Output bf16\n"
<< " 6: Input fp32, Weight fp32, Output fp32, Compute tf32)\n"
<< "arg3: tensor layout (0: Input[G, N, C, Hi, Wi], Weight[G, K, C, Y, X], Output[G, "
"N, K, Ho, Wo]\n"
<< " 1: Input[G, N, Hi, Wi, C], Weight[G, K, Y, X, C], Output[G, "
"N, Ho, Wo, K]\n"
<< " 2: Input[N, Hi, Wi, G, C], Weight[G, K, Y, X, C], Output[N, "
"Ho, Wo, G, K]\n"
<< " 3: Input[N, G, C, Hi, Wi], Weight[G, K, Y, X, C], Output[N, "
"G, K, Ho, Wo]\n"
<< " 4: Input[N, G, C, Hi, Wi], Weight[G, K, C, Y, X], Output[N, "
"G, K, Ho, Wo]\n"
<< "arg4: verification (0: no, 1: yes)\n"
<< "arg5: initialization (0: no init, 1: integer value, 2: decimal value)\n"
<< "arg6: print tensor value (0: no; 1: yes)\n"
<< "arg7: time kernel (0: no, 1: yes)\n"
<< ck::utils::conv::get_conv_param_parser_helper_msg()
<< " SplitK (-1 for internally computed split-K value, positive value to set k "
"batches explicitly, or 'all' to test all internal split-K values)\n"
<< "\nOptional arguments:\n"
<< " --instance <id> Run only the specified instance (0-indexed among valid "
"instances)\n"
<< " --list-instances List all valid instances without running\n"
<< std::endl;
}
void print_bwd_weight_instances(ConvDataType data_type,
ConvLayout layout,
ck::index_t num_dim_spatial)
{
auto print_available_instances = [&](auto num_dim_spatial_tmp,
auto in_layout,
auto wei_layout,
auto out_layout,
auto in_type,
auto wei_type,
auto out_type,
auto compute_type_a,
auto compute_type_b) {
constexpr ck::index_t NDimSpatial = num_dim_spatial_tmp.value;
using InLayout = decltype(in_layout);
using WeiLayout = decltype(wei_layout);
using OutLayout = decltype(out_layout);
using InDataType = decltype(in_type);
using WeiDataType = decltype(wei_type);
using OutDataType = decltype(out_type);
using ComputeTypeA = decltype(compute_type_a);
using ComputeTypeB = decltype(compute_type_b);
using PassThrough = ck::tensor_operation::element_wise::PassThrough;
ck::profiler::bwd_weight::print_instances<NDimSpatial,
InLayout,
WeiLayout,
OutLayout,
InDataType,
WeiDataType,
OutDataType,
PassThrough,
PassThrough,
PassThrough,
ComputeTypeA,
ComputeTypeB>();
};
constexpr auto I1 = ck::Number<1>{};
constexpr auto I2 = ck::Number<2>{};
constexpr auto I3 = ck::Number<3>{};
using F32 = float;
using F16 = ck::half_t;
using BF16 = ck::bhalf_t;
using F8 = ck::f8_t;
using BF8 = ck::bf8_t;
using TF32 = ck::tf32_t;
using namespace ck::tensor_layout::convolution;
if(num_dim_spatial == 1 && layout == ConvLayout::GNHWC_GKYXC_GNHWK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I1, GNWC{}, GKXC{}, GNWK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I1, GNWC{}, GKXC{}, GNWK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return print_available_instances(
I1, GNWC{}, GKXC{}, GNWK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I1, GNWC{}, GKXC{}, GNWK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 2 && layout == ConvLayout::GNHWC_GKYXC_GNHWK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I2, GNHWC{}, GKYXC{}, GNHWK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I2, GNHWC{}, GKYXC{}, GNHWK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return print_available_instances(
I2, GNHWC{}, GKYXC{}, GNHWK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I2, GNHWC{}, GKYXC{}, GNHWK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 2 && layout == ConvLayout::NHWGC_GKYXC_NHWGK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I2, NHWGC{}, GKYXC{}, NHWGK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I2, NHWGC{}, GKYXC{}, NHWGK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return print_available_instances(
I2, NHWGC{}, GKYXC{}, NHWGK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return print_available_instances(
I2, NHWGC{}, GKYXC{}, NHWGK{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I2, NHWGC{}, GKYXC{}, NHWGK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
else if(num_dim_spatial == 2 && layout == ConvLayout::NGCHW_GKYXC_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I2, NGCHW{}, GKYXC{}, NGKHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return print_available_instances(
I2, NGCHW{}, GKYXC{}, NGKHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
}
else if(num_dim_spatial == 2 && layout == ConvLayout::NGCHW_GKCYX_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I2, NGCHW{}, GKCYX{}, NGKHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return print_available_instances(
I2, NGCHW{}, GKCYX{}, NGKHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I2, NGCHW{}, GKCYX{}, NGKHW{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I2, NGCHW{}, GKCYX{}, NGKHW{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 3 && layout == ConvLayout::GNHWC_GKYXC_GNHWK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return print_available_instances(
I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::I8_I8_I8)
{
return print_available_instances(
I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, int8_t{}, int8_t{}, int8_t{}, int8_t{}, int8_t{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 3 && layout == ConvLayout::NHWGC_GKYXC_NHWGK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::F16_F16_F16_BF8_F8)
{
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F16{}, F16{}, F16{}, BF8{}, F8{});
}
else if(data_type == ConvDataType::I8_I8_I8)
{
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, int8_t{}, int8_t{}, int8_t{}, int8_t{}, int8_t{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
else if(num_dim_spatial == 3 && layout == ConvLayout::NGCHW_GKYXC_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I3, NGCDHW{}, GKZYXC{}, NGKDHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return print_available_instances(
I3, NGCDHW{}, GKZYXC{}, NGKDHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
}
else if(num_dim_spatial == 3 && layout == ConvLayout::NGCHW_GKCYX_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return print_available_instances(
I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return print_available_instances(
I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::F32_F32_F32)
{
return print_available_instances(
I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return print_available_instances(
I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
std::cout << "[CK_PROFILER] This data_type & layout is not implemented." << std::endl;
}
} // namespace
int profile_grouped_conv_bwd_weight(int argc, char* argv[])
{
if(argc == 6 && std::string(argv[5]) == "--instances")
{
const auto data_type = static_cast<ConvDataType>(std::stoi(argv[2]));
const auto layout = static_cast<ConvLayout>(std::stoi(argv[3]));
const ck::index_t num_dim_spatial = static_cast<ck::index_t>(std::stoi(argv[4]));
print_bwd_weight_instances(data_type, layout, num_dim_spatial);
return 0;
}
// Parse optional named arguments first
ck::index_t instance_index = -1;
bool list_instances = false;
ck::profiler::parse_named_args(argc, argv, instance_index, list_instances);
const int named_arg_count = ck::profiler::count_named_args(argc, argv);
// Adjust argc for positional argument checking
const int positional_argc = argc - named_arg_count;
// 8 for control, 1 for num_dim_spatial
if(positional_argc < 9)
{
print_helper_msg();
return 1;
}
const auto data_type = static_cast<ConvDataType>(std::stoi(argv[2]));
const auto layout = static_cast<ConvLayout>(std::stoi(argv[3]));
const bool do_verification = std::stoi(argv[4]);
const int init_method = std::stoi(argv[5]);
const bool do_log = std::stoi(argv[6]);
const bool time_kernel = std::stoi(argv[7]);
const int num_dim_spatial = std::stoi(argv[8]);
// 8 for control, 1 for num_dim_spatial, 4 for G/N/K/C, and 6 * num_dim_spatial, 1 for split-K
if(positional_argc != 8 + 1 + 4 + 6 * num_dim_spatial + 1)
{
print_helper_msg();
return 1;
}
const auto params = ck::utils::conv::parse_conv_param(num_dim_spatial, 9, argv);
const auto& split_k = std::string(argv[8 + 1 + 4 + 6 * num_dim_spatial]);
using F32 = float;
using F16 = ck::half_t;
using BF16 = ck::bhalf_t;
using F8 = ck::f8_t;
using BF8 = ck::bf8_t;
using TF32 = ck::tf32_t;
using namespace ck::tensor_layout::convolution;
constexpr auto I1 = ck::Number<1>{};
constexpr auto I2 = ck::Number<2>{};
constexpr auto I3 = ck::Number<3>{};
auto profile = [&](auto num_dim_spatial_tmp,
auto in_layout,
auto wei_layout,
auto out_layout,
auto in_type,
auto wei_type,
auto out_type,
auto compute_type_a,
auto compute_type_b) {
constexpr ck::index_t NDimSpatial = num_dim_spatial_tmp.value;
using InLayout = decltype(in_layout);
using WeiLayout = decltype(wei_layout);
using OutLayout = decltype(out_layout);
using InDataType = decltype(in_type);
using WeiDataType = decltype(wei_type);
using OutDataType = decltype(out_type);
using ComputeTypeA = decltype(compute_type_a);
using ComputeTypeB = decltype(compute_type_b);
bool pass =
ck::profiler::profile_grouped_conv_bwd_weight_impl<NDimSpatial,
InLayout,
WeiLayout,
OutLayout,
InDataType,
WeiDataType,
OutDataType,
ComputeTypeA,
ComputeTypeB>(do_verification,
init_method,
do_log,
time_kernel,
params,
split_k,
instance_index,
list_instances);
return pass ? 0 : 1;
};
if(num_dim_spatial == 1 && layout == ConvLayout::GNHWC_GKYXC_GNHWK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I1, GNWC{}, GKXC{}, GNWK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I1, GNWC{}, GKXC{}, GNWK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return profile(I1, GNWC{}, GKXC{}, GNWK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I1, GNWC{}, GKXC{}, GNWK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 2 && layout == ConvLayout::GNHWC_GKYXC_GNHWK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I2, GNHWC{}, GKYXC{}, GNHWK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I2, GNHWC{}, GKYXC{}, GNHWK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return profile(I2, GNHWC{}, GKYXC{}, GNHWK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I2, GNHWC{}, GKYXC{}, GNHWK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 2 && layout == ConvLayout::NHWGC_GKYXC_NHWGK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I2, NHWGC{}, GKYXC{}, NHWGK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I2, NHWGC{}, GKYXC{}, NHWGK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return profile(I2, NHWGC{}, GKYXC{}, NHWGK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return profile(I2, NHWGC{}, GKYXC{}, NHWGK{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I2, NHWGC{}, GKYXC{}, NHWGK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
else if(num_dim_spatial == 2 && layout == ConvLayout::NGCHW_GKYXC_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I2, NGCHW{}, GKYXC{}, NGKHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return profile(I2, NGCHW{}, GKYXC{}, NGKHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
}
else if(num_dim_spatial == 2 && layout == ConvLayout::NGCHW_GKCYX_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I2, NGCHW{}, GKCYX{}, NGKHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return profile(I2, NGCHW{}, GKCYX{}, NGKHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I2, NGCHW{}, GKCYX{}, NGKHW{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I2, NGCHW{}, GKCYX{}, NGKHW{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 3 && layout == ConvLayout::GNHWC_GKYXC_GNHWK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return profile(I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
else if(data_type == ConvDataType::I8_I8_I8)
{
return profile(
I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, int8_t{}, int8_t{}, int8_t{}, int8_t{}, int8_t{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I3, GNDHWC{}, GKZYXC{}, GNDHWK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
if(num_dim_spatial == 3 && layout == ConvLayout::NHWGC_GKYXC_NHWGK)
{
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_F32_BF16)
{
// fp32 atomic add is used for weight tensor in bf16 kernel
return profile(I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, BF16{}, F32{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return profile(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::F16_F16_F16_BF8_F8)
{
return profile(I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F16{}, F16{}, F16{}, BF8{}, F8{});
}
else if(data_type == ConvDataType::I8_I8_I8)
{
return profile(
I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, int8_t{}, int8_t{}, int8_t{}, int8_t{}, int8_t{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I3, NDHWGC{}, GKZYXC{}, NDHWGK{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
else if(num_dim_spatial == 3 && layout == ConvLayout::NGCHW_GKYXC_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I3, NGCDHW{}, GKZYXC{}, NGKDHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return profile(
I3, NGCDHW{}, GKZYXC{}, NGKDHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
}
else if(num_dim_spatial == 3 && layout == ConvLayout::NGCHW_GKCYX_NGKHW)
{
if(data_type == ConvDataType::F16_F16_F16)
{
return profile(I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, F16{}, F16{}, F16{}, F16{}, F16{});
}
if(data_type == ConvDataType::BF16_BF16_BF16)
{
return profile(
I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, BF16{}, BF16{}, BF16{}, BF16{}, BF16{});
}
if(data_type == ConvDataType::F32_F32_F32)
{
return profile(I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, F32{}, F32{}, F32{}, F32{}, F32{});
}
else if(data_type == ConvDataType::F32_F32_F32_TF32)
{
return profile(I3, NGCDHW{}, GKCZYX{}, NGKDHW{}, F32{}, F32{}, F32{}, TF32{}, TF32{});
}
}
std::cout << "this data_type & layout is not implemented" << std::endl;
return 1;
}
REGISTER_PROFILER_OPERATION(OP_NAME, OP_DESC, profile_grouped_conv_bwd_weight);
Reference in New Issue View Git Blame Copy Permalink