Skip to content
GitLab
Menu
Projects
Groups
Snippets
Loading...
Help
Help
Support
Community forum
Keyboard shortcuts
?
Submit feedback
Contribute to GitLab
Sign in / Register
Toggle navigation
Menu
Open sidebar
OpenDAS
ColossalAI
Commits
546a3744
Commit
546a3744
authored
Oct 10, 2023
by
zhuwenwen
Browse files
update readme for dtk23.10
parent
6ef178d0
Changes
1
Hide whitespace changes
Inline
Side-by-side
Showing
1 changed file
with
31 additions
and
26 deletions
+31
-26
README.md
README.md
+31
-26
No files found.
README.md
View file @
546a3744
# Colossal-AI
# <div align="center"><strong>Colossal-AI</strong></div>
## 简介
Colossal-AI提供了一系列并行组件,主要提供以下功能:提供用户友好的工具,支持编写分布式深度学习模型,只需几行即可启动分布式训练和推理。Colossal-AI官方github地址:
[
https://github.com/hpcaitech/ColossalAI
](
https://github.com/hpcaitech/ColossalAI
)
## 安装
Colossal-AI支持
...
...
@@ -7,56 +9,59 @@ Colossal-AI支持
+
Python 3.9.
### 使用pip安装
colossalai whl包下载目录:
[
https://cancon.hpccube.com:65024/4/main/colossalai/dtk23.0
4
](
https://cancon.hpccube.com:65024/4/main/colossalai/dtk23.0
4
)
colossalai whl包下载目录:
[
https://cancon.hpccube.com:65024/4/main/colossalai/dtk23.
1
0
](
https://cancon.hpccube.com:65024/4/main/colossalai/dtk23.
1
0
)
根据对应的pytorch版本和python版本,下载对应colossalai的whl包
```
shell
pip
install
colossalai
*
(
下载的colossalai的whl包
)
```
### 使用源码安装
### 使用源码
编译方式
安装
#### 编译环境准备
pytorch whl包下载目录:
[
https://cancon.hpccube.com:65024/4/main/pytorch/dtk23.04
](
https://cancon.hpccube.com:65024/4/main/pytorch/dtk23.04
)
根据python版本,下载对应pytorch的whl包
```
shell
pip
install
torch
*
(
下载的torch的whl包
)
```
提供2种环境准备方式:
1.
基于光源pytorch基础镜像环境:镜像下载地址:
[
https://sourcefind.cn/#/image/dcu/pytorch
](
https://sourcefind.cn/#/image/dcu/pytorch
)
,根据pytorch、python、dtk及系统下载对应的镜像版本。
2.
基于现有python环境:安装pytorch,pytorch whl包下载目录:
[
https://cancon.hpccube.com:65024/4/main/pytorch/dtk23.10
](
https://cancon.hpccube.com:65024/4/main/pytorch/dtk23.10
)
,根据python、dtk版本,下载对应pytorch的whl包。安装命令如下:
```
shell
pip
install
torch
*
(
下载的torch的whl包
)
pip
install
setuptools wheel
```
#### 编译安装
####
源码
编译安装
```
shell
git clone https://developer.hpccube.com/codes/aicomponent/colossalai
# 根据需要的分支进行切换
cd
colossalai
```
1、 编译whl包并安装
-
提供2种源码编译方式(进入colossalai目录):
```
1. 编译whl包并安装
python setup.py bdist_wheel
pip install dist/colossalai*
2
、 直接
编译安装
python setup.py
install
2
. 源码
编译安装
python
3
setup.py install
```
## 安装包命名
例: colossalai-0.1.13+git43ff1d4.abi0.dtk2304.torch1.10-cp37-cp37m-manylinux2014_x86_64.whl
-
colossalai: 安装包名称;
-
0.1.13: 安装包版本号;
-
git43ff1d4: git号;
-
abi0: 对应centos和rocky多系统的标识,取值为abi0(centos7),abi1(rocky8),可使用命令行动态查询获取辅助判断(echo '#include
<string>
' | gcc -x c++ -E -dM - | fgrep _GLIBCXX_USE_CXX11_ABI);
-
dtk2304: 对应dtk大版本号,命令行动态查询取值(/opt/dtk-23.04/.info/rocm_version);
-
torch1.10: 对应构建的torch版本号,命令行动态查询取值(torch.
\_\_
version__);
-
cp37-cp37m: 对应python版本号;
-
manylinux2014_x86_64: 系统架构;
## 版本号查询
-
colossalai.
\_\_
version__:与官方版本同步,查询该安装包的版本号,例如0.1.13;
-
python -c "import colossalai; print(colossalai.
\_\_
version__)",版本号与官方版本同步,查询该软件的版本号,例如0.1.13;
## Known Issue
-
无
## Note
+
若使用 pip install 下载安装过慢,可添加源:-i https://pypi.tuna.tsinghua.edu.cn/simple/
## 参考
-
[
README_ORIGIN
](
README_ORIGIN.md
)
-
[
README_zh-Hans
](
README_zh-Hans.md
)
\ No newline at end of file
-
[
README_zh-Hans
](
README_zh-Hans.md
)
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
.
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment