Skip to content

GitLab

"wrappers/python/vscode:/vscode.git/clone" did not exist on "8df547627f3a6b0d5e57f9f524a54bc4c1405689"
Commit 457308e3 authored by Jun Liu's avatar Jun Liu
Browse files

Merge branch 'develop' into amd-develop

parents a5fd9747 341ad956
Hide whitespace changes
Inline Side-by-side
Showing with 75 additions and 38 deletions
example/26_contraction/contraction_scale_xdl_fp32.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include <iostream> #include <iostream>
#include <numeric> #include <numeric>
......
example/26_contraction/contraction_scale_xdl_fp64.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include <iostream> #include <iostream>
#include <numeric> #include <numeric>
......
example/27_layernorm/common.hpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#pragma once #pragma once
......
example/27_layernorm/layernorm_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common.hpp" #include "common.hpp"
......
example/27_layernorm/layernorm_splitk_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common.hpp" #include "common.hpp"
......
example/27_layernorm/run_layernorm_example.inc
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#pragma once #pragma once
......
example/28_grouped_gemm_bias_e_permute/grouped_gemm_bias_e_permute_xdl_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include <iostream> #include <iostream>
#include <numeric> #include <numeric>
......
example/29_batched_gemm_bias_e_permute/batched_gemm_bias_e_permute_wmma_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include <iostream> #include <iostream>
#include <numeric> #include <numeric>
......
example/29_batched_gemm_bias_e_permute/batched_gemm_bias_e_permute_xdl_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include <iostream> #include <iostream>
#include <numeric> #include <numeric>
......
example/30_grouped_conv_fwd_multiple_d/CMakeLists.txt
View file @ 457308e3
add_custom_target(example_grouped_conv_fwd_multiple_d) list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
list(APPEND gpu_list2 gfx1100 gfx1101 gfx1102)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp16 grouped_conv_fwd_bias_relu_add_xdl_fp16.cpp) set(target 0)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp32 grouped_conv_fwd_bias_relu_add_xdl_fp32.cpp) foreach(gpu IN LISTS GPU_TARGETS)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_bf16 grouped_conv_fwd_bias_relu_add_xdl_bf16.cpp) if(gpu IN_LIST gpu_list1 AND target EQUAL 0)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int8 grouped_conv_fwd_bias_relu_add_xdl_int8.cpp) add_custom_target(example_grouped_conv_fwd_multiple_d)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp16) add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp16 grouped_conv_fwd_bias_relu_add_xdl_fp16.cpp)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp32) add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_fp32 grouped_conv_fwd_bias_relu_add_xdl_fp32.cpp)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_bf16) add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_bf16 grouped_conv_fwd_bias_relu_add_xdl_bf16.cpp)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int8) add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int8 grouped_conv_fwd_bias_relu_add_xdl_int8.cpp)
if(USE_BITINT_EXTENSION_INT4) add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp16)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int4 grouped_conv_fwd_bias_relu_add_xdl_int4.cpp) add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_fp32)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_bf16)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int8)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int4) if(USE_BITINT_EXTENSION_INT4)
endif() # USE_BITINT_EXTENSION_INT4 add_example_executable(example_grouped_conv_fwd_bias_relu_add_xdl_int4 grouped_conv_fwd_bias_relu_add_xdl_int4.cpp)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_bias_relu_add_xdl_int4)
endif() # USE_BITINT_EXTENSION_INT4
add_example_executable(example_grouped_conv_fwd_xdl_fp16 grouped_conv_fwd_xdl_fp16.cpp)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_xdl_fp16)
set(target 1)
endif()
endforeach()
if(GPU_TARGETS MATCHES "gfx1100" OR GPU_TARGETS MATCHES "gfx1101" OR GPU_TARGETS MATCHES "gfx1102") set(target 0)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_fp16 grouped_conv_fwd_bias_relu_add_wmma_fp16.cpp) foreach(gpu IN LISTS GPU_TARGETS)
endif() if(gpu IN_LIST gpu_list2 AND target EQUAL 0)
add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_fp16 grouped_conv_fwd_bias_relu_add_wmma_fp16.cpp)
add_example_executable(example_grouped_conv_fwd_xdl_fp16 grouped_conv_fwd_xdl_fp16.cpp) add_example_executable(example_grouped_conv_fwd_bias_relu_add_wmma_int8 grouped_conv_fwd_bias_relu_add_wmma_int8.cpp)
set(target 1)
add_dependencies(example_grouped_conv_fwd_multiple_d example_grouped_conv_fwd_xdl_fp16) endif()
endforeach()
example/30_grouped_conv_fwd_multiple_d/common.hpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#pragma once #pragma once
......
example/30_grouped_conv_fwd_multiple_d/common_wmma.hpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#pragma once #pragma once
......
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_wmma_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common_wmma.hpp" #include "common_wmma.hpp"
......
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_wmma_int8.cpp 0 → 100644
View file @ 457308e3
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
#include "common_wmma.hpp"
// kernel data types
using InKernelDataType = I8;
using WeiKernelDataType = I8;
using AccDataType = I32;
using CShuffleDataType = I8;
using BiasKernelDataType = I8;
using ResidualKernelDataType = I8;
using OutKernelDataType = I8;
// tensor data types
using InUserDataType = InKernelDataType;
using WeiUserDataType = WeiKernelDataType;
using OutUserDataType = OutKernelDataType;
using InElementOp = PassThrough;
using WeiElementOp = PassThrough;
using OutElementOp = ck::tensor_operation::element_wise::AddReluAdd;
#include "run_grouped_conv_fwd_bias_relu_add_wmma_example.inc"
int main(int argc, char* argv[]) { return !run_grouped_conv_fwd_bias_relu_add_example(argc, argv); }
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_xdl_bf16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common.hpp" #include "common.hpp"
......
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_xdl_fp16.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common.hpp" #include "common.hpp"
......
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_xdl_fp32.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common.hpp" #include "common.hpp"
......
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_xdl_int4.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#ifndef CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4 #ifndef CK_EXPERIMENTAL_BIT_INT_EXTENSION_INT4
#error Should compile this file with ck::int4_t support #error Should compile this file with ck::int4_t support
......
example/30_grouped_conv_fwd_multiple_d/grouped_conv_fwd_bias_relu_add_xdl_int8.cpp
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
#include "common.hpp" #include "common.hpp"
......
example/30_grouped_conv_fwd_multiple_d/run_grouped_conv_fwd_bias_relu_add_example.inc
View file @ 457308e3
// SPDX-License-Identifier: MIT // SPDX-License-Identifier: MIT
// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved. // Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
template <typename BiasLay, typename ResidualLay> template <typename BiasLay, typename ResidualLay>
struct LayoutSetting struct LayoutSetting
......
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