mirror of
https://github.com/ROCm/composable_kernel.git
synced 2026-04-19 22:39:03 +00:00
3048028897
* Move kernel implementation files under impl directory. * Update examples paths. * Update device kernel impl include paths. * Update tensor operation instances include paths. * Update profiler and tests include paths. * Clang-format * Update include paths for batched gemm reduce * Refactor UnitTest ConvNDBwdWeight. * Refactor fwd and bwd data convND UT. * Fix used test macro. * Fix include path. * Fix include paths. * Fix include paths in profiler and tests. * Fix include paths. Co-authored-by: Adam Osewski <aosewski@amd.com>
46 lines
4.0 KiB
C++
46 lines
4.0 KiB
C++
// SPDX-License-Identifier: MIT
|
|
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
|
|
|
|
#ifndef CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
|
|
#error Should compile this file with ck::int4_t support
|
|
#endif
|
|
|
|
#include "common.hpp"
|
|
|
|
#include "ck/tensor_operation/gpu/device/impl/device_gemm_dl.hpp"
|
|
|
|
using ADataType = ck::int4_t;
|
|
using BDataType = ck::int4_t;
|
|
using CDataType = ck::int4_t;
|
|
using KernelADataType = int8_t;
|
|
using KernelBDataType = int8_t;
|
|
using KernelCDataType = int8_t;
|
|
using AccDataType = int32_t;
|
|
|
|
using ALayout = Col;
|
|
using BLayout = Row;
|
|
using CLayout = Row;
|
|
|
|
using AElementOp = PassThrough;
|
|
using BElementOp = PassThrough;
|
|
using CElementOp = PassThrough;
|
|
|
|
static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
|
|
|
|
// clang-format off
|
|
using DeviceGemmInstance = ck::tensor_operation::device::DeviceGemmDl
|
|
// ######| AData| BData| CData| AccData| ALayout| BLayout| CLayout| A| B| C| GEMM| Block| MPer| NPer| K0Per| K1| M1Per| N1Per| KPer| M11N11Thread| M11N11Thread| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| BBlockTransfer| BBlockTransfer| BBlockTransfer| BBlockTransfer| BBlockTransfer| BBlockTransfer| BBlockTransfer| CThreadTransfer| CThreadTransfer| CThreadTransfer|
|
|
// ######| Type| Type| Type| Type| | | | Elementwise| Elementwise| Elementwise| Spacialization| Size| Block| Block| Block| | ThreadM111| ThreadN111| Thread| ClusterM110Xs| ClusterN110Xs| ThreadSliceLengths| ThreadClusterLengths| ThreadCluster| SrcAccess| SrcVectorTensor| SrcVectorTensor| DstVectorTensor| ThreadSliceLengths| ThreadClusterLengths| ThreadCluster| SrcAccess| SrcVectorTensor| SrcVectorTensor| DstVectorTensor| SrcDstAccess| SrcDstVectorDim| DstScalarPerVector|
|
|
// ######| | | | | | | | Operation| Operation| Operation| | | | | | | | | | | | K0_M0_M1_K1| K0_M0_M1_K1| ArrangeOrder| Order| Lengths_K0_M0_M1_K1| ContiguousDimOrder| Lengths_K0_M0_M1_K1| K0_N0_N1_K1| K0_N0_N1_K1| ArrangeOrder| Order| Lengths_K0_N0_N1_K1| ContiguousDimOrder| Lengths_K0_N0_N1_K1| Order| | |
|
|
// ######| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
|
|
< KernelADataType, KernelBDataType, KernelCDataType, AccDataType, ALayout, BLayout, CLayout, AElementOp, BElementOp, CElementOp, GemmDefault, 256, 128, 128, 16, 4, 4, 4, 1, S<8, 2>, S<8, 2>, S<2, 1, 4, 4>, S<8, 1, 32, 1>, S<0, 3, 1, 2>, S<0, 3, 1, 2>, S<1, 1, 4, 1>, S<0, 3, 1, 2>, S<1, 1, 4, 4>, S<2, 1, 4, 4>, S<8, 1, 32, 1>, S<0, 3, 1, 2>, S<0, 3, 1, 2>, S<1, 1, 4, 1>, S<0, 3, 1, 2>, S<1, 1, 4, 4>, S<0, 1, 2, 3, 4, 5>, 5, 4>;
|
|
// clang-format on
|
|
|
|
using ReferenceGemmInstance = ck::tensor_operation::host::
|
|
ReferenceGemm<ADataType, BDataType, CDataType, AccDataType, AElementOp, BElementOp, CElementOp>;
|
|
|
|
#define BUILD_INT4_EXAMPLE
|
|
#include "run_gemm_example.inc"
|
|
|
|
int main(int argc, char* argv[]) { return !run_gemm_example(argc, argv); }
|