Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
magicwindyyd
mindspore
提交
e4517964
M
mindspore
项目概览
magicwindyyd
/
mindspore
与 Fork 源项目一致
Fork自
MindSpore / mindspore
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
M
mindspore
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
提交
e4517964
编写于
4月 29, 2020
作者:
M
meixiaowei
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
support relative and full paths
上级
8ab5e011
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
32 addition
and
10 deletion
+32
-10
example/resnet101_imagenet2012/run_distribute_train.sh
example/resnet101_imagenet2012/run_distribute_train.sh
+19
-7
example/resnet101_imagenet2012/run_standalone_train.sh
example/resnet101_imagenet2012/run_standalone_train.sh
+13
-3
未找到文件。
example/resnet101_imagenet2012/run_distribute_train.sh
浏览文件 @
e4517964
...
...
@@ -20,23 +20,35 @@ then
exit
1
fi
if
[
!
-f
$1
]
get_real_path
(){
if
[
"
${
1
:0:1
}
"
==
"/"
]
;
then
echo
"
$1
"
else
echo
"
$(
realpath
-m
$PWD
/
$1
)
"
fi
}
PATH1
=
$(
get_real_path
$1
)
PATH2
=
$(
get_real_path
$2
)
echo
$PATH1
echo
$PATH2
if
[
!
-f
$PATH1
]
then
echo
"error:
DMINDSPORE_HCCL_CONFIG_PATH=
$
1
is not a file"
echo
"error:
MINDSPORE_HCCL_CONFIG_PATH=
$PATH
1
is not a file"
exit
1
fi
if
[
!
-d
$2
]
if
[
!
-d
$
PATH
2
]
then
echo
"error: DATASET_PATH=
$2
is not a directory"
echo
"error: DATASET_PATH=
$
PATH
2
is not a directory"
exit
1
fi
ulimit
-u
unlimited
export
DEVICE_NUM
=
8
export
RANK_SIZE
=
8
export
MINDSPORE_HCCL_CONFIG_PATH
=
$1
export
RANK_TABLE_FILE
=
$1
export
MINDSPORE_HCCL_CONFIG_PATH
=
$
PATH
1
export
RANK_TABLE_FILE
=
$
PATH
1
for
((
i
=
0
;
i<
${
DEVICE_NUM
}
;
i++
))
do
...
...
@@ -49,6 +61,6 @@ do
cd
./train_parallel
$i
||
exit
echo
"start training for rank
$RANK_ID
, device
$DEVICE_ID
"
env
>
env.log
python train.py
--do_train
=
True
--run_distribute
=
True
--device_num
=
$DEVICE_NUM
--dataset_path
=
$2
&> log &
python train.py
--do_train
=
True
--run_distribute
=
True
--device_num
=
$DEVICE_NUM
--dataset_path
=
$
PATH
2
&> log &
cd
..
done
example/resnet101_imagenet2012/run_standalone_train.sh
浏览文件 @
e4517964
...
...
@@ -20,9 +20,19 @@ then
exit
1
fi
if
[
!
-d
$1
]
get_real_path
(){
if
[
"
${
1
:0:1
}
"
==
"/"
]
;
then
echo
"
$1
"
else
echo
"
$(
realpath
-m
$PWD
/
$1
)
"
fi
}
PATH1
=
$(
get_real_path
$1
)
echo
$PATH1
if
[
!
-d
$PATH1
]
then
echo
"error: DATASET_PATH=
$1
is not a directory"
echo
"error: DATASET_PATH=
$
PATH
1
is not a directory"
exit
1
fi
...
...
@@ -42,5 +52,5 @@ cp *.sh ./train
cd
./train
||
exit
echo
"start training for device
$DEVICE_ID
"
env
>
env.log
python train.py
--do_train
=
True
--dataset_path
=
$1
&> log &
python train.py
--do_train
=
True
--dataset_path
=
$
PATH
1
&> log &
cd
..
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录