Skip to content

GitLab

Commit 6ac1d6a2 authored by illsilin's avatar illsilin
Browse files

merging from public repo

parents e60c5aea 42fc8edd
Hide whitespace changes
Inline Side-by-side
Showing with 141 additions and 85 deletions
.github/CODEOWNERS
View file @ 6ac1d6a2
* @zjing14 @asroy @junliume @illsilin @carlushuang @aosewski
* @zjing14 @junliume @illsilin @carlushuang @aosewski
# Documentation files
docs/* @saadrahim @LisaDelaney
*.md @saadrahim @LisaDelaney
*.rst @saadrahim @LisaDelaney
# Header directory
library/include/* @saadrahim @LisaDelaney
docs/* @ROCm/rocm-documentation
*.md @ROCm/rocm-documentation
*.rst @ROCm/rocm-documentation
# Header directory for Doxygen documentation
library/include/* @ROCm/rocm-documentation
Dockerfile
View file @ 6ac1d6a2
......@@ -44,7 +44,6 @@ ENV PATH=$PATH:${SCCACHE_INSTALL_LOCATION}
RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-unauthenticated \
build-essential \
cmake \
ccache \
git \
hip-rocclr \
iputils-ping \
......@@ -74,6 +73,10 @@ RUN apt-get update && DEBIAN_FRONTEND=noninteractive apt-get install -y --allow-
apt-get clean && \
rm -rf /var/lib/apt/lists/*
#Install latest ccache
RUN git clone https://github.com/ccache/ccache.git && \
cd ccache && mkdir build && cd build && cmake .. && make install
#Install ninja build tracing tools
RUN wget -qO /usr/local/bin/ninja.gz https://github.com/ninja-build/ninja/releases/latest/download/ninja-linux.zip
RUN gunzip /usr/local/bin/ninja.gz
......@@ -111,7 +114,7 @@ ENV LANG=C.UTF-8
RUN groupadd -f render
# Install the new rocm-cmake version
RUN git clone -b master https://github.com/RadeonOpenCompute/rocm-cmake.git && \
RUN git clone -b master https://github.com/ROCm/rocm-cmake.git && \
cd rocm-cmake && mkdir build && cd build && \
cmake .. && cmake --build . && cmake --build . --target install
......@@ -123,7 +126,7 @@ RUN sh -c "echo compiler version = '$compiler_version'"
RUN sh -c "echo compiler commit = '$compiler_commit'"
RUN if ( [ "$compiler_version" = "amd-staging" ] || [ "$compiler_version" = "amd-mainline-open" ] ) && [ "$compiler_commit" = "" ]; then \
git clone -b "$compiler_version" https://github.com/RadeonOpenCompute/llvm-project.git && \
git clone -b "$compiler_version" https://github.com/ROCm/llvm-project.git && \
cd llvm-project && mkdir build && cd build && \
cmake -DCMAKE_INSTALL_PREFIX=/opt/rocm/llvm -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=1 -DLLVM_TARGETS_TO_BUILD="AMDGPU;X86" -DLLVM_ENABLE_PROJECTS="clang;lld" -DLLVM_ENABLE_RUNTIMES="compiler-rt" ../llvm && \
make -j 8 ; \
......@@ -131,7 +134,7 @@ RUN if ( [ "$compiler_version" = "amd-staging" ] || [ "$compiler_version" = "amd
fi
RUN if ( [ "$compiler_version" = "amd-staging" ] || [ "$compiler_version" = "amd-mainline-open" ] ) && [ "$compiler_commit" != "" ]; then \
git clone -b "$compiler_version" https://github.com/RadeonOpenCompute/llvm-project.git && \
git clone -b "$compiler_version" https://github.com/ROCm/llvm-project.git && \
cd llvm-project && git checkout "$compiler_commit" && echo "checking out commit $compiler_commit" && mkdir build && cd build && \
cmake -DCMAKE_INSTALL_PREFIX=/opt/rocm/llvm -DCMAKE_BUILD_TYPE=Release -DLLVM_ENABLE_ASSERTIONS=1 -DLLVM_TARGETS_TO_BUILD="AMDGPU;X86" -DLLVM_ENABLE_PROJECTS="clang;lld" -DLLVM_ENABLE_RUNTIMES="compiler-rt" ../llvm && \
make -j 8 ; \
......
Jenkinsfile
View file @ 6ac1d6a2
def rocmnode(name) {
return '(rocmtest || miopen) && ' + name
return '(rocmtest || miopen) && (' + name + ')'
}
def show_node_info() {
......@@ -7,6 +7,7 @@ def show_node_info() {
echo "NODE_NAME = \$NODE_NAME"
lsb_release -sd
uname -r
cat /sys/module/amdgpu/version
ls /opt/ -la
"""
}
......@@ -33,6 +34,10 @@ def runShell(String command){
def getDockerImageName(){
def img
if (params.USE_CUSTOM_DOCKER != ""){
img = "${params.USE_CUSTOM_DOCKER}"
}
else{
if (params.ROCMVERSION != "6.0.1"){
if (params.COMPILER_VERSION == "") {
img = "${env.CK_DOCKERHUB}:ck_ub20.04_rocm${params.ROCMVERSION}"
......@@ -61,6 +66,7 @@ def getDockerImageName(){
}
}
}
}
return img
}
......@@ -134,7 +140,9 @@ def buildDocker(install_prefix){
//force building the new docker if that parameter is true
echo "Building image: ${image_name}"
retimage = docker.build("${image_name}", dockerArgs + ' .')
retimage.push()
withDockerRegistry([ credentialsId: "docker_test_cred", url: "" ]) {
retimage.push()
}
sh 'docker images -q -f dangling=true | xargs --no-run-if-empty docker rmi'
}
else{
......@@ -146,7 +154,9 @@ def buildDocker(install_prefix){
catch(Exception ex){
echo "Unable to locate image: ${image_name}. Building image now"
retimage = docker.build("${image_name}", dockerArgs + ' .')
retimage.push()
withDockerRegistry([ credentialsId: "docker_test_cred", url: "" ]) {
retimage.push()
}
}
}
......@@ -254,18 +264,24 @@ def cmake_build(Map conf=[:]){
""")
sh cmd3
}
def setup_cmd = conf.get("setup_cmd", "${cmake_envs} cmake ${setup_args} .. ")
// reduce parallelism when compiling, clang uses too much memory
def nt = nthreads()
def build_cmd = conf.get("build_cmd", "${build_envs} dumb-init make -j${nt} ${config_targets}")
def cmd
def execute_cmd = conf.get("execute_cmd", "")
def cmd = conf.get("cmd", """
if(!setup_args.contains("NO_CK_BUILD")){
def setup_cmd = conf.get("setup_cmd", "${cmake_envs} cmake ${setup_args} .. ")
def build_cmd = conf.get("build_cmd", "${build_envs} dumb-init make -j${nt} ${config_targets}")
cmd = conf.get("cmd", """
${setup_cmd}
${build_cmd}
${execute_cmd}
""")
}
else{
cmd = conf.get("cmd", """
${execute_cmd}
""")
}
echo cmd
......@@ -361,8 +377,8 @@ def runCKProfiler(Map conf=[:]){
(retimage, image) = getDockerImage(conf)
withDockerContainer(image: image, args: dockerOpts) {
timeout(time: 5, unit: 'MINUTES'){
sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
sh 'rocminfo | tee rocminfo.log'
if ( !runShell('grep -n "gfx" rocminfo.log') ){
throw new Exception ("GPU not found")
}
else{
......@@ -375,20 +391,6 @@ def runCKProfiler(Map conf=[:]){
echo "The job was cancelled or aborted"
throw e
}
catch(Exception ex) {
retimage = docker.build("${image}", dockerArgs + " --no-cache .")
withDockerContainer(image: image, args: dockerOpts) {
timeout(time: 5, unit: 'MINUTES'){
sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
throw new Exception ("GPU not found")
}
else{
echo "GPU is OK"
}
}
}
}
withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
timeout(time: 24, unit: 'HOURS')
......@@ -469,6 +471,7 @@ def Build_CK(Map conf=[:]){
show_node_info()
env.HSA_ENABLE_SDMA=0
env.DOCKER_BUILDKIT=1
checkout scm
def image = getDockerImageName()
......@@ -483,25 +486,35 @@ def Build_CK(Map conf=[:]){
if (params.COMPILER_VERSION == "amd-staging" || params.COMPILER_VERSION == "amd-mainline-open" || params.COMPILER_COMMIT != ""){
dockerOpts = dockerOpts + " --env HIP_CLANG_PATH='/llvm-project/build/bin' "
}
def video_id = sh(returnStdout: true, script: 'getent group video | cut -d: -f3')
def render_id = sh(returnStdout: true, script: 'getent group render | cut -d: -f3')
dockerOpts = dockerOpts + " --group-add=${video_id} --group-add=${render_id} "
echo "Docker flags: ${dockerOpts}"
def variant = env.STAGE_NAME
def retimage
def navi_node = 0
def mi300_node = 0
gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel-internal') {
gitStatusWrapper(credentialsId: "${env.status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel-internal') {
try {
(retimage, image) = getDockerImage(conf)
withDockerContainer(image: image, args: dockerOpts) {
timeout(time: 5, unit: 'MINUTES'){
sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo | tee clinfo.log'
if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
sh 'rocminfo | tee rocminfo.log'
if ( !runShell('grep -n "gfx" rocminfo.log') ){
throw new Exception ("GPU not found")
}
else{
echo "GPU is OK"
}
if ( runShell('grep -n "gfx1030" clinfo.log') || runShell('grep -n "gfx1101" clinfo.log') ){
if ( runShell('grep -n "gfx1030" rocminfo.log') || runShell('grep -n "gfx1101" rocminfo.log') ){
navi_node = 1
echo "This is a Navi node"
}
if ( runShell('grep -n "gfx942" rocminfo.log') ){
mi300_node = 1
echo "This is MI300 node"
}
}
}
......@@ -510,23 +523,6 @@ def Build_CK(Map conf=[:]){
echo "The job was cancelled or aborted"
throw e
}
catch(Exception ex) {
retimage = docker.build("${image}", dockerArgs + " --no-cache .")
withDockerContainer(image: image, args: dockerOpts) {
timeout(time: 5, unit: 'MINUTES'){
sh 'PATH="/opt/rocm/opencl/bin:/opt/rocm/opencl/bin/x86_64:$PATH" clinfo |tee clinfo.log'
if ( runShell('grep -n "Number of devices:.*. 0" clinfo.log') ){
throw new Exception ("GPU not found")
}
else{
echo "GPU is OK"
}
if ( runShell('grep -n "gfx1030" clinfo.log') || runShell('grep -n "gfx1101" clinfo.log') ){
navi_node = 1
}
}
}
}
withDockerContainer(image: image, args: dockerOpts + ' -v=/var/jenkins/:/var/jenkins') {
timeout(time: 24, unit: 'HOURS')
{
......@@ -540,8 +536,8 @@ def Build_CK(Map conf=[:]){
sh 'tar -zcvf ckProfiler.tar.gz bin/ckProfiler'
stash "ckProfiler.tar.gz"
}
if (params.RUN_FULL_QA){
// build deb packages
if (params.RUN_FULL_QA && mi300_node == 0 ){
// build deb packages for all MI100/200/300 targets and prepare to export
sh 'make -j package'
archiveArtifacts artifacts: 'composablekernel-ckprofiler_*.deb'
archiveArtifacts artifacts: 'composablekernel-tests_*.deb'
......@@ -606,7 +602,7 @@ def process_results(Map conf=[:]){
def variant = env.STAGE_NAME
def retimage
gitStatusWrapper(credentialsId: "${status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel-internal') {
gitStatusWrapper(credentialsId: "${env.status_wrapper_creds}", gitHubContext: "Jenkins - ${variant}", account: 'ROCm', repo: 'composable_kernel-internal') {
try {
(retimage, image) = getDockerImage(conf)
}
......@@ -665,6 +661,10 @@ pipeline {
name: "BUILD_DOCKER",
defaultValue: false,
description: "Force building docker image (default: false), set to true if docker image needs to be updated.")
string(
name: 'USE_CUSTOM_DOCKER',
defaultValue: '',
description: 'If you want to use a custom docker image, please specify it here (default: leave blank).')
string(
name: 'ROCMVERSION',
defaultValue: '6.0',
......@@ -709,6 +709,10 @@ pipeline {
name: "RUN_PERFORMANCE_TESTS",
defaultValue: false,
description: "Run the performance tests (default: OFF)")
booleanParam(
name: "RUN_CODEGEN_TESTS",
defaultValue: true,
description: "Run the codegen tests (default: ON)")
}
environment{
dbuser = "${dbuser}"
......@@ -787,7 +791,34 @@ pipeline {
}
}
}
stage("Run Codegen Tests")
{
parallel
{
stage("Run Codegen Tests on MI100/MI200")
{
when {
beforeAgent true
expression { params.RUN_CODEGEN_TESTS.toBoolean() }
}
options { retry(2) }
agent{ label rocmnode("gfx908 || gfx90a")}
environment{
setup_args = "NO_CK_BUILD"
execute_args = """ cd ../codegen && rm -rf build && mkdir build && cd build && \
cmake -D CMAKE_PREFIX_PATH=/opt/rocm \
-D CMAKE_CXX_COMPILER=/opt/rocm/llvm/bin/clang++ \
-D CMAKE_BUILD_TYPE=Release \
-D GPU_TARGETS="gfx908;gfx90a" \
-DCMAKE_CXX_FLAGS=" -O3 " .. && make -j check"""
}
steps{
buildHipClangJobAndReboot(setup_args:setup_args, no_reboot:true, build_type: 'Release', execute_cmd: execute_args)
cleanWs()
}
}
}
}
stage("Build CK and run Tests")
{
parallel
......@@ -815,6 +846,26 @@ pipeline {
cleanWs()
}
}
stage("Build CK and run Tests on MI300")
{
when {
beforeAgent true
expression { params.RUN_FULL_QA.toBoolean() }
}
agent{ label rocmnode("gfx942") }
environment{
setup_args = """ -DCMAKE_INSTALL_PREFIX=../install -DGPU_TARGETS="gfx942" -DCMAKE_CXX_FLAGS=" -O3 " """
execute_args = """ cd ../client_example && rm -rf build && mkdir build && cd build && \
cmake -DCMAKE_PREFIX_PATH="${env.WORKSPACE}/install;/opt/rocm" \
-DGPU_TARGETS="gfx942" \
-DCMAKE_CXX_COMPILER="${build_compiler()}" \
-DCMAKE_CXX_FLAGS=" -O3 " .. && make -j """
}
steps{
Build_CK_and_Reboot(setup_args: setup_args, config_targets: "install", no_reboot:true, build_type: 'Release', execute_cmd: execute_args, prefixpath: '/usr/local')
cleanWs()
}
}
stage("Build CK and run Tests on MI100/MI200")
{
when {
......
client_example/01_gemm/gemm.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -83,7 +83,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/02_gemm_add_add_fastgelu/gemm_add_add_fastgelu.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -92,7 +92,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/02_gemm_add_add_fastgelu/gemm_add_add_fastgelu_generic.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -93,7 +93,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/02_gemm_add_add_fastgelu/gemm_add_fastgelu.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -88,7 +88,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/02_gemm_add_add_fastgelu/gemm_add_fastgelu_generic.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -89,7 +89,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/02_gemm_add_add_fastgelu/gemm_fastgelu.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -84,7 +84,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/02_gemm_add_add_fastgelu/gemm_fastgelu_generic.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -85,7 +85,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/03_gemm_layernorm/gemm_add_add_layernorm_naive.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......@@ -17,6 +17,8 @@
using F16 = ck::half_t;
using F32 = float;
using Row = ck::tensor_layout::gemm::RowMajor;
using ADataType = F16;
using BDataType = F16;
using BiasDataType = F32;
......@@ -191,7 +193,7 @@ int main()
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/03_gemm_layernorm/gemm_add_relu_add_layernorm_welford.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <iostream>
......@@ -78,7 +78,7 @@ int main(int argc, char* argv[])
[](std::size_t nRow, std::size_t nCol, std::size_t stride, auto layout) {
using Layout = decltype(layout);
if constexpr(std::is_same<Layout, ck::tensor_layout::gemm::RowMajor>::value)
if constexpr(std::is_same<Layout, Row>::value)
{
return (nRow - 1) * stride + nCol;
}
......
client_example/04_contraction/contraction_bilinear_fp32.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <numeric>
......
client_example/04_contraction/contraction_bilinear_fp64.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <numeric>
......
client_example/04_contraction/contraction_g1m2n3k1_add_xdl_fp16.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <numeric>
......
client_example/04_contraction/contraction_scale_fp32.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <numeric>
......
client_example/04_contraction/contraction_scale_fp64.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <numeric>
......
client_example/05_layernorm/layernorm2d_bwd_data.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......
client_example/05_layernorm/layernorm2d_bwd_gamma_beta.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......
client_example/05_layernorm/layernorm2d_fwd.cpp
View file @ 6ac1d6a2
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#include <iomanip>
#include <vector>
......
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