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
7dd18c0d
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
7dd18c0d
编写于
8月 27, 2020
作者:
M
meng_chunyang
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
update the copy rules for fp16 models
上级
34cdab58
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
13 addition
and
14 deletion
+13
-14
mindspore/lite/test/models_fp16.cfg
mindspore/lite/test/models_fp16.cfg
+3
-3
mindspore/lite/test/run_benchmark_nets.sh
mindspore/lite/test/run_benchmark_nets.sh
+10
-11
未找到文件。
mindspore/lite/test/models_fp16.cfg
浏览文件 @
7dd18c0d
detect-deeper-halfdeeper-mbv1-shortcut-400-400_nopostprocess_simplified.fp16
detect-deeper-halfdeeper-mbv1-shortcut-400-400_nopostprocess_simplified.fp16
model_emotions_0727_nosoftmax.fp16
model_emotions_0727_nosoftmax.
tflite.
fp16
mtk_isface.fp16
mtk_isface.fp16
mtk_landmark.fp16
mtk_landmark.fp16
mtk_pose_tuku.fp16
mtk_pose_tuku.fp16
mtk_age_gender.fp16
mtk_age_gender.
tflite.
fp16
mtk_model_face_dress.fp16
mtk_model_face_dress.
tflite.
fp16
mindspore/lite/test/run_benchmark_nets.sh
浏览文件 @
7dd18c0d
...
@@ -403,17 +403,6 @@ rm -rf ${basepath}/ms_models
...
@@ -403,17 +403,6 @@ rm -rf ${basepath}/ms_models
mkdir
-p
${
basepath
}
/ms_models
mkdir
-p
${
basepath
}
/ms_models
ms_models_path
=
${
basepath
}
/ms_models
ms_models_path
=
${
basepath
}
/ms_models
echo
"start convert models ... :"
# Copy fp16 ms models:
while
read
line
;
do
model_name
=
${
line
}
if
[[
$model_name
==
\#
*
]]
;
then
continue
fi
echo
"cp '
${
models_path
}
'/'
${
model_name
}
'.ms'
${
ms_models_path
}
'/'
${
model_name
}
'.ms"
>>
"
${
run_benchmark_log_file
}
"
cp
$models_path
/
${
model_name
}
.ms
${
ms_models_path
}
/
${
model_name
}
.ms
done
<
${
models_fp16_config
}
# Convert tflite models:
# Convert tflite models:
while
read
line
;
do
while
read
line
;
do
model_name
=
${
line
}
model_name
=
${
line
}
...
@@ -483,6 +472,16 @@ while read line; do
...
@@ -483,6 +472,16 @@ while read line; do
./converter_lite
--fmk
=
TFLITE
--modelFile
=
${
models_path
}
/
${
model_name
}
--outputFile
=
${
ms_models_path
}
/
${
model_name
}
--quantType
=
AwareTraining
||
Convert_status
=
$?
./converter_lite
--fmk
=
TFLITE
--modelFile
=
${
models_path
}
/
${
model_name
}
--outputFile
=
${
ms_models_path
}
/
${
model_name
}
--quantType
=
AwareTraining
||
Convert_status
=
$?
done
<
${
models_tflite_awaretraining_config
}
done
<
${
models_tflite_awaretraining_config
}
# Copy fp16 ms models:
while
read
line
;
do
model_name
=
${
line
%.*
}
if
[[
$model_name
==
\#
*
]]
;
then
continue
fi
echo
'cp '
${
ms_models_path
}
'/'
${
model_name
}
'.ms'
${
ms_models_path
}
'/'
${
model_name
}
'.fp16.ms'
cp
${
ms_models_path
}
/
${
model_name
}
.ms
${
ms_models_path
}
/
${
model_name
}
.fp16.ms
done
<
${
models_fp16_config
}
# Check all result and return value
# Check all result and return value
if
[[
${
Convert_status
}
=
0
]]
;
then
if
[[
${
Convert_status
}
=
0
]]
;
then
echo
"convert is ended"
echo
"convert is ended"
...
...
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