Skip to content

GitLab

Commit 1dbdab56 authored by Jing Zhang's avatar Jing Zhang
Browse files

merge develop

parents d2e49b23 bac7df8f
Hide whitespace changes
Inline Side-by-side
Showing with 250 additions and 204 deletions
library/src/tensor_operation_instance/gpu/batched_gemm/CMakeLists.txt
View file @ 1dbdab56
#device_batched_gemm_instance add_instance_library(device_batched_gemm_instance
set(DEVICE_BATCHED_GEMM_INSTANCE_SOURCE device_batched_gemm_xdl_f16_f16_f16_gmk_gkn_gmn_instance.cpp
device_batched_gemm_xdl_f16_f16_f16_gmk_gkn_gmn_instance.cpp; device_batched_gemm_xdl_f16_f16_f16_gmk_gnk_gmn_instance.cpp
device_batched_gemm_xdl_f16_f16_f16_gmk_gnk_gmn_instance.cpp; device_batched_gemm_xdl_f16_f16_f16_gkm_gkn_gmn_instance.cpp
device_batched_gemm_xdl_f16_f16_f16_gkm_gkn_gmn_instance.cpp; device_batched_gemm_xdl_f16_f16_f16_gkm_gnk_gmn_instance.cpp
device_batched_gemm_xdl_f16_f16_f16_gkm_gnk_gmn_instance.cpp; device_batched_gemm_xdl_bf16_bf16_bf16_gmk_gkn_gmn_instance.cpp
device_batched_gemm_xdl_bf16_bf16_bf16_gmk_gkn_gmn_instance.cpp; device_batched_gemm_xdl_bf16_bf16_bf16_gmk_gnk_gmn_instance.cpp
device_batched_gemm_xdl_bf16_bf16_bf16_gmk_gnk_gmn_instance.cpp; device_batched_gemm_xdl_bf16_bf16_bf16_gkm_gkn_gmn_instance.cpp
device_batched_gemm_xdl_bf16_bf16_bf16_gkm_gkn_gmn_instance.cpp; device_batched_gemm_xdl_bf16_bf16_bf16_gkm_gnk_gmn_instance.cpp
device_batched_gemm_xdl_bf16_bf16_bf16_gkm_gnk_gmn_instance.cpp; device_batched_gemm_xdl_f32_f32_f32_gmk_gkn_gmn_instance.cpp
device_batched_gemm_xdl_f32_f32_f32_gmk_gkn_gmn_instance.cpp; device_batched_gemm_xdl_f32_f32_f32_gmk_gnk_gmn_instance.cpp
device_batched_gemm_xdl_f32_f32_f32_gmk_gnk_gmn_instance.cpp; device_batched_gemm_xdl_f32_f32_f32_gkm_gkn_gmn_instance.cpp
device_batched_gemm_xdl_f32_f32_f32_gkm_gkn_gmn_instance.cpp; device_batched_gemm_xdl_f32_f32_f32_gkm_gnk_gmn_instance.cpp
device_batched_gemm_xdl_f32_f32_f32_gkm_gnk_gmn_instance.cpp; device_batched_gemm_xdl_int8_int8_int8_gmk_gkn_gmn_instance.cpp
device_batched_gemm_xdl_int8_int8_int8_gmk_gkn_gmn_instance.cpp; device_batched_gemm_xdl_int8_int8_int8_gmk_gnk_gmn_instance.cpp
device_batched_gemm_xdl_int8_int8_int8_gmk_gnk_gmn_instance.cpp; device_batched_gemm_xdl_int8_int8_int8_gkm_gkn_gmn_instance.cpp
device_batched_gemm_xdl_int8_int8_int8_gkm_gkn_gmn_instance.cpp; device_batched_gemm_xdl_int8_int8_int8_gkm_gnk_gmn_instance.cpp
device_batched_gemm_xdl_int8_int8_int8_gkm_gnk_gmn_instance.cpp;
) )
add_library(device_batched_gemm_instance OBJECT ${DEVICE_BATCHED_GEMM_INSTANCE_SOURCE})
# target_compile_features(device_batched_gemm_instance PUBLIC)
set_target_properties(device_batched_gemm_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
# install(TARGETS device_batched_gemm_instance LIBRARY DESTINATION lib)
clang_tidy_check(device_batched_gemm_instance)
library/src/tensor_operation_instance/gpu/batched_gemm_gemm/CMakeLists.txt 0 → 100644
View file @ 1dbdab56
add_instance_library(device_batched_gemm_gemm_instance
device_batched_gemm_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
)
library/src/tensor_operation_instance/gpu/batched_gemm_gemm/device_batched_gemm_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp 0 → 100644
View file @ 1dbdab56
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
#include <cstdlib>
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
#include "ck/tensor_operation/gpu/device/device_batched_gemm_gemm_xdl_cshuffle.hpp"
#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
namespace ck {
namespace tensor_operation {
namespace device {
namespace instance {
using F16 = ck::half_t;
using F32 = float;
using Row = ck::tensor_layout::gemm::RowMajor;
using Col = ck::tensor_layout::gemm::ColumnMajor;
template <ck::index_t... Is>
using S = ck::Sequence<Is...>;
using PassThrough = ck::tensor_operation::element_wise::PassThrough;
static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
// c[g, m, n] = a[g, m, k] * b[g, n, k]
using device_batched_gemm_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instances = std::tuple<
// clang-format off
//################################| ALayout| B0Layout| B1Layout| CLayout| AData| B0Data| B1Data| CData| AccData| CShuffle| A| B0| Acc0| B1| C| GEMM| NumGemmK| Block| Gemm01| Gemm0| Gemm0| Gemm1| Gemm1| AK1| BK1| B1K1| MPer| NPer| Gemm0| Gemm0| Gemm1| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockLds| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockLds| CShuffle| CShuffle| CBlockTransferClusterLengths| CBlockTransfer|
//################################| | | | | Type| Type| Type| Type| Type| DataType| Elementwise| Elementwise| Elementwise| Elementwise| Elementwise| Specialization| Prefetch| Size| MPer| NPer| KPer| NPer| KPer| | | | XDL| XDL| MXdl| NXdl| NXdl| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraM| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraN| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MWaveMPerXdl| ScalarPerVector|
//################################| | | | | | | | | | | Operation| Operation| Operation| Operation| Operation| | Stage| | Block| Block| Block| Block| Block| | | | | | Per| Per| Per| Lengths_K0_M_K1| ArrangeOrder| | | PerVector| PerVector_K1| | Lengths_K0_N_K1| ArrangeOrder| | | PerVector| PerVector_K1| | Lengths_K0_N_K1| ArrangeOrder| | | PerVector| PerVector_K1| | PerShuffle| PerShuffle| _NBlock_NWaveNPerXdl| _NWaveNPerXdl|
//################################| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Wave| Wave| Wave| | | | | | | | | | | | | | | | | | | | | | | | | |
DeviceBatchedGemmGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 256, 128, 32, 128, 32, 8, 8, 2, 32, 32, 2, 4, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S< 8, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>,
DeviceBatchedGemmGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 128, 128, 32, 128, 32, 8, 8, 2, 32, 32, 1, 4, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S< 8, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>,
DeviceBatchedGemmGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 128, 128, 32, 64, 32, 8, 8, 2, 32, 32, 1, 4, 2, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<16, 16, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>,
DeviceBatchedGemmGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 128, 64, 32, 128, 32, 8, 8, 2, 32, 32, 1, 2, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S< 8, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>
// clang-format on
>;
void add_device_batched_gemm_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance(
std::vector<std::unique_ptr<DeviceBatchedGemmGemm<Row,
Col,
Row,
Row,
F16,
F16,
F16,
F16,
PassThrough,
PassThrough,
PassThrough,
PassThrough,
PassThrough>>>& instances)
{
add_device_operation_instances(
instances,
device_batched_gemm_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instances{});
}
} // namespace instance
} // namespace device
} // namespace tensor_operation
} // namespace ck
library/src/tensor_operation_instance/gpu/batched_gemm_reduce/CMakeLists.txt
View file @ 1dbdab56
set(DEVICE_BATCHED_GEMM_REDUCE_INSTANCE_SOURCE add_instance_library(device_batched_gemm_reduce_instance
device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gmk_gkn_gmn_instance.cpp device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gmk_gkn_gmn_instance.cpp
device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gmk_gnk_gmn_instance.cpp device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gmk_gnk_gmn_instance.cpp
device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gkm_gkn_gmn_instance.cpp device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gkm_gkn_gmn_instance.cpp
device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gkm_gnk_gmn_instance.cpp device_batched_gemm_reduce_xdl_cshuffle_f16_f16_f16_f32_f32_gkm_gnk_gmn_instance.cpp
) )
add_instance_library(device_batched_gemm_reduce_instance OBJECT ${DEVICE_BATCHED_GEMM_REDUCE_INSTANCE_SOURCE})
target_compile_features(device_batched_gemm_reduce_instance PUBLIC)
set_target_properties(device_batched_gemm_reduce_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_batched_gemm_reduce_instance)
library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm/CMakeLists.txt 0 → 100644
View file @ 1dbdab56
add_instance_library(device_batched_gemm_softmax_gemm_instance
device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp
)
library/src/tensor_operation_instance/gpu/batched_gemm_softmax_gemm/device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance.cpp 0 → 100644
View file @ 1dbdab56
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
#include <cstdlib>
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
#include "ck/tensor_operation/gpu/device/device_batched_gemm_softmax_gemm_xdl_cshuffle.hpp"
#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
namespace ck {
namespace tensor_operation {
namespace device {
namespace instance {
using F16 = ck::half_t;
using F32 = float;
using Row = ck::tensor_layout::gemm::RowMajor;
using Col = ck::tensor_layout::gemm::ColumnMajor;
template <ck::index_t... Is>
using S = ck::Sequence<Is...>;
using PassThrough = ck::tensor_operation::element_wise::PassThrough;
static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
// c[g, m, n] = a[g, m, k] * b[g, n, k]
using device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instances =
std::tuple<
// clang-format off
//#######################################| ALayout| B0Layout| B1Layout| CLayout| AData| B0Data| B1Data| CData| AccData| CShuffle| A| B0| Acc0| B1| C| GEMM| NumGemmK| Block| Gemm01| Gemm0| Gemm0| Gemm1| Gemm1| AK1| BK1| B1K1| MPer| NPer| Gemm0| Gemm0| Gemm1| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockTransfer| B0BlockLds| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockTransfer| B1BlockLds| CShuffle| CShuffle| CBlockTransferClusterLengths| CBlockTransfer|
//#######################################| | | | | Type| Type| Type| Type| Type| DataType| Elementwise| Elementwise| Elementwise| Elementwise| Elementwise| Specialization| Prefetch| Size| MPer| NPer| KPer| NPer| KPer| | | | XDL| XDL| MXdl| NXdl| NXdl| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraM| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraN| ThreadCluster| ThreadCluster| SrcAccessOrder| SrcVectorDim| SrcScalar| DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave| _MBlock_MWaveMPerXdl| ScalarPerVector|
//#######################################| | | | | | | | | | | Operation| Operation| Operation| Operation| Operation| | Stage| | Block| Block| Block| Block| Block| | | | | | Per| Per| Per| Lengths_K0_M_K1| ArrangeOrder| | | PerVector| PerVector_K1| | Lengths_K0_N_K1| ArrangeOrder| | | PerVector| PerVector_K1| | Lengths_K0_N_K1| ArrangeOrder| | | PerVector| PerVector_K1| | PerShuffle| PerShuffle| _NBlock_NWaveNPerXdl| _NWaveNPerXdl|
//#######################################| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Wave| Wave| Wave| | | | | | | | | | | | | | | | | | | | | | | | | |
DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 256, 128, 32, 128, 32, 8, 8, 2, 32, 32, 2, 4, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S< 8, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>,
DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 128, 128, 32, 128, 32, 8, 8, 2, 32, 32, 1, 4, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S< 8, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>,
DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 128, 128, 32, 64, 32, 8, 8, 2, 32, 32, 1, 4, 2, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<16, 16, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>,
DeviceBatchedGemmSoftmaxGemm_Xdl_CShuffle< Row, Col, Row, Row, F16, F16, F16, F16, F32, F16, PassThrough, PassThrough, PassThrough, PassThrough, PassThrough, GemmDefault, 1, 256, 128, 64, 32, 128, 32, 8, 8, 2, 32, 32, 1, 2, 4, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S<4, 64, 1>, S<1, 0, 2>, S<1, 0, 2>, 2, 8, 8, true, S< 8, 32, 1>, S<0, 2, 1>, S<0, 2, 1>, 1, 4, 2, false, 1, 2, S<1, 32, 1, 8>, 8>
// clang-format on
>;
void add_device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instance(
std::vector<std::unique_ptr<DeviceBatchedGemmSoftmaxGemm<Row,
Col,
Row,
Row,
F16,
F16,
F16,
F16,
PassThrough,
PassThrough,
PassThrough,
PassThrough,
PassThrough>>>& instances)
{
add_device_operation_instances(
instances,
device_batched_gemm_softmax_gemm_xdl_cshuffle_f16_f16_f16_f16_gmk_gnk_gno_gmo_instances{});
}
} // namespace instance
} // namespace device
} // namespace tensor_operation
} // namespace ck
library/src/tensor_operation_instance/gpu/contraction_bilinear/CMakeLists.txt
View file @ 1dbdab56
# device_contraction_bilinear_instance add_instance_library(device_contraction_bilinear_instance
set(DEVICE_CONTRACTION_BILINEAR_INSTANCE_SOURCE
device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_kknn_instance.cpp device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_kknn_instance.cpp
device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_knnn_instance.cpp device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_knnn_instance.cpp
device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_mknn_instance.cpp device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_mknn_instance.cpp
device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_mnnn_instance.cpp device_contraction_bilinear_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_f32_mnnn_instance.cpp
) )
add_library(device_contraction_bilinear_instance OBJECT ${DEVICE_CONTRACTION_BILINEAR_INSTANCE_SOURCE})
set_target_properties(device_contraction_bilinear_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_contraction_bilinear_instance)
library/src/tensor_operation_instance/gpu/contraction_scale/CMakeLists.txt
View file @ 1dbdab56
# device_contraction_scale_instance add_instance_library(device_contraction_scale_instance
set(DEVICE_CONTRACTION_SCALE_INSTANCE_SOURCE
device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_kkn_instance.cpp device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_kkn_instance.cpp
device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_knn_instance.cpp device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_knn_instance.cpp
device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_mkn_instance.cpp device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_mkn_instance.cpp
device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_mnn_instance.cpp device_contraction_scale_m2_n2_k2_xdl_c_shuffle_f32_f32_f32_mnn_instance.cpp
) )
add_library(device_contraction_scale_instance OBJECT ${DEVICE_CONTRACTION_SCALE_INSTANCE_SOURCE})
set_target_properties(device_contraction_scale_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_contraction_scale_instance)
library/src/tensor_operation_instance/gpu/conv1d_bwd_data/CMakeLists.txt
View file @ 1dbdab56
# device_conv1d_bwd_data_instance add_instance_library(device_conv1d_bwd_data_instance
set(DEVICE_CONV1D_BWD_DATA_INSTANCE_SOURCE device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f16_instance.cpp
device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f16_instance.cpp; device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f32_instance.cpp
device_conv1d_bwd_data_xdl_nwc_kxc_nwk_f32_instance.cpp; device_conv1d_bwd_data_xdl_nwc_kxc_nwk_bf16_instance.cpp
device_conv1d_bwd_data_xdl_nwc_kxc_nwk_bf16_instance.cpp; device_conv1d_bwd_data_xdl_nwc_kxc_nwk_int8_instance.cpp
device_conv1d_bwd_data_xdl_nwc_kxc_nwk_int8_instance.cpp;
) )
add_library(device_conv1d_bwd_data_instance OBJECT ${DEVICE_CONV1D_BWD_DATA_INSTANCE_SOURCE})
target_compile_features(device_conv1d_bwd_data_instance PUBLIC)
set_target_properties(device_conv1d_bwd_data_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
rocm_install(TARGETS device_conv1d_bwd_data_instance)
clang_tidy_check(device_conv1d_bwd_data_instance)
library/src/tensor_operation_instance/gpu/conv1d_bwd_weight/CMakeLists.txt
View file @ 1dbdab56
#device_conv1d_bwd_weight_instance add_instance_library(device_conv1d_bwd_weight_instance
set(DEVICE_CONV1D_BWD_WEIGHT_INSTANCE_SOURCE device_conv1d_bwd_weight_xdl_nwc_kxc_nwk_f16_instance.cpp
device_conv1d_bwd_weight_xdl_nwc_kxc_nwk_f16_instance.cpp; device_conv1d_bwd_weight_xdl_nwc_kxc_nwk_f32_instance.cpp
device_conv1d_bwd_weight_xdl_nwc_kxc_nwk_f32_instance.cpp; device_conv1d_bwd_weight_xdl_nwc_kxc_nwk_bf16_instance.cpp
device_conv1d_bwd_weight_xdl_nwc_kxc_nwk_bf16_instance.cpp;
) )
add_library(device_conv1d_bwd_weight_instance OBJECT ${DEVICE_CONV1D_BWD_WEIGHT_INSTANCE_SOURCE})
target_compile_features(device_conv1d_bwd_weight_instance PUBLIC)
set_target_properties(device_conv1d_bwd_weight_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
rocm_install(TARGETS device_conv1d_bwd_weight_instance)
clang_tidy_check(device_conv1d_bwd_weight_instance)
library/src/tensor_operation_instance/gpu/conv2d_bwd_data/CMakeLists.txt
View file @ 1dbdab56
# device_conv2d_bwd_data_instance add_instance_library(device_conv2d_bwd_data_instance
set(DEVICE_CONV2D_BWD_DATA_INSTANCE_SOURCE device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f32_instance.cpp; device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_f16_instance.cpp; device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp; device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_int8_instance.cpp
device_conv2d_bwd_data_xdl_nhwc_kyxc_nhwk_int8_instance.cpp;
) )
add_library(device_conv2d_bwd_data_instance OBJECT ${DEVICE_CONV2D_BWD_DATA_INSTANCE_SOURCE})
set_target_properties(device_conv2d_bwd_data_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_conv2d_bwd_data_instance)
library/src/tensor_operation_instance/gpu/conv2d_bwd_weight/CMakeLists.txt
View file @ 1dbdab56
#device_conv2d_bwd_weight_instance add_instance_library(device_conv2d_bwd_weight_instance
set(DEVICE_CONV2D_BWD_WEIGHT_INSTANCE_SOURCE device_conv2d_bwd_weight_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
device_conv2d_bwd_weight_xdl_nhwc_kyxc_nhwk_f16_instance.cpp; device_conv2d_bwd_weight_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
device_conv2d_bwd_weight_xdl_nhwc_kyxc_nhwk_f32_instance.cpp; device_conv2d_bwd_weight_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
device_conv2d_bwd_weight_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp;
) )
add_library(device_conv2d_bwd_weight_instance OBJECT ${DEVICE_CONV2D_BWD_WEIGHT_INSTANCE_SOURCE})
target_compile_features(device_conv2d_bwd_weight_instance PUBLIC)
set_target_properties(device_conv2d_bwd_weight_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
rocm_install(TARGETS device_conv2d_bwd_weight_instance)
clang_tidy_check(device_conv2d_bwd_weight_instance)
library/src/tensor_operation_instance/gpu/conv2d_fwd/CMakeLists.txt
View file @ 1dbdab56
# device_conv2d_fwd_instance add_instance_library(device_conv2d_fwd_instance
set(DEVICE_CONV2D_FWD_INSTANCE_SOURCE device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp
device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f32_instance.cpp; device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp
device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_f16_instance.cpp; device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp
device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_bf16_instance.cpp; device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp
device_conv2d_fwd_xdl_nhwc_kyxc_nhwk_int8_instance.cpp; device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp
device_conv2d_fwd_xdl_c_shuffle_nhwc_kyxc_nhwk_f16_instance.cpp;
) )
add_library(device_conv2d_fwd_instance OBJECT ${DEVICE_CONV2D_FWD_INSTANCE_SOURCE})
set_target_properties(device_conv2d_fwd_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_conv2d_fwd_instance)
library/src/tensor_operation_instance/gpu/conv2d_fwd_bias_relu/CMakeLists.txt
View file @ 1dbdab56
# device_conv2d_fwd_bias_relu_instance add_instance_library(device_conv2d_fwd_bias_relu_instance
set(DEVICE_CONV2D_FWD_BIAS_RELU_INSTANCE_SOURCE device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instance.cpp
device_conv2d_fwd_xdl_c_shuffle_bias_relu_nhwc_kyxc_nhwk_f16_instance.cpp;
) )
add_library(device_conv2d_fwd_bias_relu_instance OBJECT ${DEVICE_CONV2D_FWD_BIAS_RELU_INSTANCE_SOURCE})
set_target_properties(device_conv2d_fwd_bias_relu_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_conv2d_fwd_bias_relu_instance)
library/src/tensor_operation_instance/gpu/conv2d_fwd_bias_relu_add/CMakeLists.txt
View file @ 1dbdab56
# device_conv2d_fwd_bias_relu_add_instance add_instance_library(device_conv2d_fwd_bias_relu_add_instance
set(DEVICE_CONV2D_FWD_BIAS_RELU_ADD_INSTANCE_SOURCE device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instance.cpp
device_conv2d_fwd_xdl_c_shuffle_bias_relu_add_nhwc_kyxc_nhwk_f16_instance.cpp;
) )
add_library(device_conv2d_fwd_bias_relu_add_instance OBJECT ${DEVICE_CONV2D_FWD_BIAS_RELU_ADD_INSTANCE_SOURCE})
set_target_properties(device_conv2d_fwd_bias_relu_add_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_conv2d_fwd_bias_relu_add_instance)
library/src/tensor_operation_instance/gpu/conv3d_bwd_data/CMakeLists.txt
View file @ 1dbdab56
# device_conv3d_bwd_data_instance add_instance_library(device_conv3d_bwd_data_instance
set(DEVICE_CONV3D_BWD_DATA_INSTANCE_SOURCE device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f16_instance.cpp
device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f16_instance.cpp; device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f32_instance.cpp
device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_f32_instance.cpp; device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_bf16_instance.cpp
device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_bf16_instance.cpp; device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_int8_instance.cpp
device_conv3d_bwd_data_xdl_ndhwc_kzyxc_ndhwk_int8_instance.cpp;
) )
add_library(device_conv3d_bwd_data_instance OBJECT ${DEVICE_CONV3D_BWD_DATA_INSTANCE_SOURCE})
target_compile_features(device_conv3d_bwd_data_instance PUBLIC)
set_target_properties(device_conv3d_bwd_data_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
rocm_install(TARGETS device_conv3d_bwd_data_instance)
clang_tidy_check(device_conv3d_bwd_data_instance)
library/src/tensor_operation_instance/gpu/conv3d_bwd_weight/CMakeLists.txt
View file @ 1dbdab56
#device_conv3d_bwd_weight_instance add_instance_library(device_conv3d_bwd_weight_instance
set(DEVICE_CONV3D_BWD_WEIGHT_INSTANCE_SOURCE device_conv3d_bwd_weight_xdl_ndhwc_kzyxc_ndhwk_f16_instance.cpp
device_conv3d_bwd_weight_xdl_ndhwc_kzyxc_ndhwk_f16_instance.cpp; device_conv3d_bwd_weight_xdl_ndhwc_kzyxc_ndhwk_f32_instance.cpp
device_conv3d_bwd_weight_xdl_ndhwc_kzyxc_ndhwk_f32_instance.cpp; device_conv3d_bwd_weight_xdl_ndhwc_kzyxc_ndhwk_bf16_instance.cpp
device_conv3d_bwd_weight_xdl_ndhwc_kzyxc_ndhwk_bf16_instance.cpp;
) )
add_library(device_conv3d_bwd_weight_instance OBJECT ${DEVICE_CONV3D_BWD_WEIGHT_INSTANCE_SOURCE})
target_compile_features(device_conv3d_bwd_weight_instance PUBLIC)
set_target_properties(device_conv3d_bwd_weight_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
rocm_install(TARGETS device_conv3d_bwd_weight_instance)
clang_tidy_check(device_conv3d_bwd_weight_instance)
library/src/tensor_operation_instance/gpu/elementwise/CMakeLists.txt
View file @ 1dbdab56
set(DEVICE_ELEMENTWISE_INSTANCE_SOURCE add_instance_library(device_elementwise_instance
device_normalize_instance.cpp device_normalize_instance.cpp
) )
add_instance_library(device_elementwise_instance ${DEVICE_ELEMENTWISE_INSTANCE_SOURCE})
target_compile_features(device_elementwise_instance PUBLIC)
set_target_properties(device_elementwise_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
clang_tidy_check(device_elementwise_instance)
library/src/tensor_operation_instance/gpu/elementwise/device_normalize_instance.cpp
View file @ 1dbdab56
...@@ -6,7 +6,7 @@ ...@@ -6,7 +6,7 @@
#include "ck/ck.hpp" #include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp" #include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp" #include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
#include "ck/tensor_operation/gpu/device/device_5ary_elementwise.hpp" #include "ck/tensor_operation/gpu/device/device_elementwise.hpp"
#include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp" #include "ck/library/tensor_operation_instance/add_device_operation_instance.hpp"
namespace ck { namespace ck {
...@@ -27,19 +27,17 @@ using outputType = F16; ...@@ -27,19 +27,17 @@ using outputType = F16;
using Normalize = ck::tensor_operation::element_wise::Normalize; using Normalize = ck::tensor_operation::element_wise::Normalize;
using device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances = std::tuple< using device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances = std::tuple<
// clang-format off // clang-format off
//###################|in | mean| square_mean| gamma| beta| out| ComputeDataType| functor| NDim| MPerThread| in, mean, square_mean, gamma, beta, out ScalarPerVector| //###################|<in, mean, square_mean, gamma, beta>| <out>| functor| NDim| MPerThread| <in, mean, square_mean, gamma, beta ScalarPerVector>| <out ScalarPerVector>|
//###################|in | mean| square_mean| gamma| beta| out| ComputeDataType| functor| NDim| MPerThread| in, mean, square_mean, gamma, beta, out ScalarPerVector| DeviceElementwise<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2, 8, Sequence<8, 1, 1, 8, 8>, Sequence<8> >,
//###################|in | mean| square_mean| gamma| beta| out| ComputeDataType| functor| NDim| MPerThread| in, mean, square_mean, gamma, beta, out ScalarPerVector| DeviceElementwise<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2, 4, Sequence<4, 1, 1, 4, 4>, Sequence<4> >,
//###################|in | mean| square_mean| gamma| beta| out| ComputeDataType| functor| NDim| MPerThread| in, mean, square_mean, gamma, beta, out ScalarPerVector| DeviceElementwise<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2, 2, Sequence<2, 1, 1, 2, 2>, Sequence<2> >,
Device5AryElementwise<F16, F32, F32, F16, F16, F16, F32, Normalize, 2, 8, 8, 1, 1, 8, 8, 8 >, DeviceElementwise<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2, 1, Sequence<1, 1, 1, 1, 1>, Sequence<1> >
Device5AryElementwise<F16, F32, F32, F16, F16, F16, F32, Normalize, 2, 4, 4, 1, 1, 4, 4, 4 >,
Device5AryElementwise<F16, F32, F32, F16, F16, F16, F32, Normalize, 2, 2, 2, 1, 1, 2, 2, 2 >,
Device5AryElementwise<F16, F32, F32, F16, F16, F16, F32, Normalize, 2, 1, 1, 1, 1, 1, 1, 1 >
// clang-format on // clang-format on
>; >;
void add_device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances( void add_device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances(
std::vector<DeviceElementwisePtr<5, 1, 2, Normalize>>& instances) std::vector<DeviceElementwiseBasePtr<Tuple<F16, F32, F32, F16, F16>, Tuple<F16>, Normalize, 2>>&
instances)
{ {
add_device_operation_instances( add_device_operation_instances(
instances, device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances{}); instances, device_normalize_from_mean_squaremean_f16_f32_f32_f16_f16_instances{});
......
library/src/tensor_operation_instance/gpu/gemm/CMakeLists.txt
View file @ 1dbdab56
set(DEVICE_GEMM_INSTANCE_SOURCE add_instance_library(device_gemm_instance
device_gemm_xdl_f64_f64_f64_mk_kn_mn_instance.cpp; device_gemm_xdl_f64_f64_f64_mk_kn_mn_instance.cpp
device_gemm_xdl_f64_f64_f64_mk_nk_mn_instance.cpp; device_gemm_xdl_f64_f64_f64_mk_nk_mn_instance.cpp
device_gemm_xdl_f64_f64_f64_km_kn_mn_instance.cpp; device_gemm_xdl_f64_f64_f64_km_kn_mn_instance.cpp
device_gemm_xdl_f64_f64_f64_km_nk_mn_instance.cpp; device_gemm_xdl_f64_f64_f64_km_nk_mn_instance.cpp
device_gemm_xdl_f32_f32_f32_mk_kn_mn_instance.cpp; device_gemm_xdl_f32_f32_f32_mk_kn_mn_instance.cpp
device_gemm_xdl_f32_f32_f32_mk_nk_mn_instance.cpp; device_gemm_xdl_f32_f32_f32_mk_nk_mn_instance.cpp
device_gemm_xdl_f32_f32_f32_km_kn_mn_instance.cpp; device_gemm_xdl_f32_f32_f32_km_kn_mn_instance.cpp
device_gemm_xdl_f32_f32_f32_km_nk_mn_instance.cpp; device_gemm_xdl_f32_f32_f32_km_nk_mn_instance.cpp
device_gemm_xdl_f16_f16_f16_mk_kn_mn_instance.cpp; device_gemm_xdl_f16_f16_f16_mk_kn_mn_instance.cpp
device_gemm_xdl_f16_f16_f16_mk_nk_mn_instance.cpp; device_gemm_xdl_f16_f16_f16_mk_nk_mn_instance.cpp
device_gemm_xdl_f16_f16_f16_km_kn_mn_instance.cpp; device_gemm_xdl_f16_f16_f16_km_kn_mn_instance.cpp
device_gemm_xdl_f16_f16_f16_km_nk_mn_instance.cpp; device_gemm_xdl_f16_f16_f16_km_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_i8_i8_i8_mk_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_i8_i8_i8_mk_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_i8_i8_i8_mk_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_i8_i8_i8_mk_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_i8_i8_i8_km_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_i8_i8_i8_km_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_i8_i8_i8_km_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_i8_i8_i8_km_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_bf16_bf16_bf16_mk_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_bf16_bf16_bf16_mk_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_bf16_bf16_bf16_mk_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_bf16_bf16_bf16_mk_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_bf16_bf16_bf16_km_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_bf16_bf16_bf16_km_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_bf16_bf16_bf16_km_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_bf16_bf16_bf16_km_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_f16_f16_f16_mk_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_f16_f16_f16_mk_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_f16_f16_f16_mk_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_f16_f16_f16_mk_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_f16_f16_f16_km_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_f16_f16_f16_km_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_f16_f16_f16_km_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_f16_f16_f16_km_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_f32_f32_f32_mk_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_f32_f32_f32_mk_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_f32_f32_f32_mk_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_f32_f32_f32_mk_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_f32_f32_f32_km_kn_mn_instance.cpp; device_gemm_xdl_c_shuffle_f32_f32_f32_km_kn_mn_instance.cpp
device_gemm_xdl_c_shuffle_f32_f32_f32_km_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_f32_f32_f32_km_nk_mn_instance.cpp
device_gemm_xdl_c_shuffle_2_stage_f16_f16_f16_mk_nk_mn_instance.cpp; device_gemm_xdl_c_shuffle_2_stage_f16_f16_f16_mk_nk_mn_instance.cpp
device_gemm_dl_f32_f32_f32_mk_kn_mn_instance.cpp; device_gemm_dl_f32_f32_f32_mk_kn_mn_instance.cpp
device_gemm_dl_f32_f32_f32_mk_nk_mn_instance.cpp; device_gemm_dl_f32_f32_f32_mk_nk_mn_instance.cpp
device_gemm_dl_f32_f32_f32_km_kn_mn_instance.cpp; device_gemm_dl_f32_f32_f32_km_kn_mn_instance.cpp
device_gemm_dl_f32_f32_f32_km_nk_mn_instance.cpp; device_gemm_dl_f32_f32_f32_km_nk_mn_instance.cpp
device_gemm_dl_f16_f16_f16_mk_kn_mn_instance.cpp; device_gemm_dl_f16_f16_f16_mk_kn_mn_instance.cpp
device_gemm_dl_f16_f16_f16_mk_nk_mn_instance.cpp; device_gemm_dl_f16_f16_f16_mk_nk_mn_instance.cpp
device_gemm_dl_f16_f16_f16_km_kn_mn_instance.cpp; device_gemm_dl_f16_f16_f16_km_kn_mn_instance.cpp
device_gemm_dl_f16_f16_f16_km_nk_mn_instance.cpp; device_gemm_dl_f16_f16_f16_km_nk_mn_instance.cpp
device_gemm_dl_i8_i8_i8_mk_kn_mn_instance.cpp; device_gemm_dl_i8_i8_i8_mk_kn_mn_instance.cpp
device_gemm_dl_i8_i8_i8_mk_nk_mn_instance.cpp; device_gemm_dl_i8_i8_i8_mk_nk_mn_instance.cpp
device_gemm_dl_i8_i8_i8_km_kn_mn_instance.cpp; device_gemm_dl_i8_i8_i8_km_kn_mn_instance.cpp
device_gemm_dl_i8_i8_i8_km_nk_mn_instance.cpp; device_gemm_dl_i8_i8_i8_km_nk_mn_instance.cpp
) )
add_library(device_gemm_instance OBJECT ${DEVICE_GEMM_INSTANCE_SOURCE})
target_compile_features(device_gemm_instance PUBLIC)
set_target_properties(device_gemm_instance PROPERTIES POSITION_INDEPENDENT_CODE ON)
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment