Huawei_Technology
/
MindSpore-AKG
Not watched
1
0
Fork 0
Code
Releases 1 Wiki evaluate Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain HPC
You can not select more than 25 topics Topics must start with a chinese character,a letter or number, can include dashes ('-') and can be up to 35 characters long.
25 Commits
2 Branches
7.6 MB
378 Download
 
 
 
 
 
 
Tree: a5a856cd2c
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'a5a856cd2c'
${ noResults }
mindspore-ci-bot a5a856cd2c !63 Update releasenote 1.2.0 5 years ago
.gitee init 5 years ago
.github init 5 years ago
cmake init 5 years ago
docs init 5 years ago
python/akg [TUNING] add the gpu-tuning process to master 5 years ago
scripts init 5 years ago
src enhance mapping: drop axes that cannot be mapped to block before block mapping 5 years ago
tests [TUNING] add the gpu-tuning process to master 5 years ago
third_party enhance mapping: drop axes that cannot be mapped to block before block mapping 5 years ago
.clang-format init 5 years ago
.gitignore init 5 years ago
CMakeLists.txt Can not set _GLIBCXX_USE_CXX11_ABI=0 in current profiling 5 years ago
CONTRIBUTING.md init 5 years ago
LICENSE init 5 years ago
NOTICE init 5 years ago
README.md init 5 years ago
README_CN.md init 5 years ago
RELEASE.md update releasenote 1.2.0 5 years ago
Third_Party_Open_Source_Software_Notice init 5 years ago
build.sh sync master 5 years ago
requirements.txt init 5 years ago
setup.py init 5 years ago

README.md

查看中文

What Is AKG

AKG(Auto Kernel Generator) is an optimizer for operators in Deep Learning Networks. It provides the ability to automatically fuse ops with specific patterns. AKG works with MindSpore-GraphKernel to improve the performance of networks running on different hardware backends.

AKG composes with three basic optimization module, normalization, auto schedule and backend optimization.

  • normalization. In order to solve the limitation in expression ability of polyhedral(which can only process static linear programs), the computation IR needs to be normalized first. The mainly optimization of normalization module includes auto-inline, loop fusing, common subexpression elimination and so on.

  • auto schedule. Base on polyhedral technology, the auto schedule module mainly have auto-vectorization, auto-tiling, thread/block mapping, dependency analysis and memory promotion.

  • backend optimization. The backend optimization module mainly consists of TensorCore acceleration, double buffer optimization, storage flatten optimization and inject sync optimization.

Hardware Backends Support

At present, GPU V100/A100 are supported. More Backends are on the list.

Build

Build With MindSpore

See MindSpore README.md for details.

Build Standalone

We suggest you build and run akg together with MindSpore. And we also provide a way to run case in standalone mode for convenience sake.
Refer to MindSpore Installation for more information about compilation dependencies.

bash build.sh -e $target // target can set 'gpu'

Run Standalone

  1. Set Environment
  • GPU V100/A100 
    cd tests
source ./test_env.sh gpu
  1. Run test
  • GPU V100/A100 
    cd tests/operators/gpu
python3 test_all.py -s "op_name" #replace op_name with the operator name which you want to test

Contributing

Welcome contributions. See MindSpore Contributor Wiki for
more details.

Release Notes

The release notes, see our RELEASE.

License

Apache License 2.0

Delete task

Are you sure you want to delete this task? Once this task is deleted, it cannot be recovered.

cancel
confirm

AKG(Auto Kernel Generator)对深度神经网络中的算子进行优化，并提供特定模式下的算子自动融合功能。AKG与MindSpore的图算融合功能协同工作，可提升在不同硬件后端上运行网络的性能。

Python C++ Text Cuda CMake other