wangwei
/
parser
Not watched
1
0
Fork 0
Code
Releases 0 Wiki Activity
Issues 0 Pull Requests 0 Datasets Model Cloudbrain
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.
759 Commits
14 Branches
86 MB
972 Download
 
 
 
 
 
 
Tree: b0846ee77c
Branches Tags
${ item.name }
Create branch ${ searchTerm }
from 'b0846ee77c'
${ noResults }
TangQunzhang b0846ee77c Fix codecheck 4 years ago
.gitee Init the project with Apache 2.0 license 6 years ago
cmake !615 删除gflags依赖 4 years ago
inc/external codecheck clean 5 years ago
metadef @ 1c8737f98c !618 update submodule 4 years ago
parser Fix codecheck 4 years ago
tests !619 增加AutoMappingSubgraphDataFormat接口设置子图上的DATA节点的format 4 years ago
.gitignore Add UT and format check 5 years ago
.gitmodules update submodule 5 years ago
CMakeLists.txt !598 0706 黄传蓝 4 years ago
LICENSE Init the project with Apache 2.0 license 6 years ago
NOTICE Init the project with Apache 2.0 license 6 years ago
OWNERS !617 update owners 4 years ago
README.md update README.md. 6 years ago
RELEASE.md update readme 6 years ago
build.sh allinone 4 years ago
classify_rule.txt add parser classify rule file 5 years ago
classify_rule.yaml add ymal 4 years ago

README.md

Ascend CANN Parser

Ascend CANN Parser（简称parser）配合TF_Adapter、 ATC工具、IR构图等使用，开发者通过以上工具，借助parser能方便地将第三方框架的算法表示转换成Ascend IR，充分利用昇腾AI处理器卓越的运算能力。
parser系统框架

安装

parser以动态库的方式被调用。

源码安装

parser支持源码编译，进行源码编译前，首先确保你有昇腾910AI处理器的环境，同时确保系统满足以下要求：

  • GCC >= 7.3.0
  • CMake >= 3.14.0
  • Autoconf >= 2.64
  • Libtool >= 2.4.6
  • Automake >= 1.15.1

下载源码

git clone --recursive 
https://gitee.com/ascend/parser.git -b development
cd parser
chmod +x build.sh

源码编译

在parser根目录执行以下命令编译：

bash build.sh

编译完成之后，相应的动态库文件会生成在output文件夹中

贡献

欢迎参与贡献。

Release Notes

Release Notes请参考RELEASE。

许可证

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

Ascend CANN Parser（简称parser）配合TF_Adapter、 ATC工具、IR构图等使用，开发者通过以上工具，借助parser能方便地将第三方框架的算法表示转换成Ascend IR，充分利用昇腾AI处理器卓越的运算能力

C++ Python Text CMake PureBasic other