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
PaddlePaddle
PaddleClas
提交
2d66aeeb
P
PaddleClas
项目概览
PaddlePaddle
/
PaddleClas
接近 2 年 前同步成功
通知
116
Star
4999
Fork
1114
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
19
列表
看板
标记
里程碑
合并请求
6
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
PaddleClas
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
19
Issue
19
列表
看板
标记
里程碑
合并请求
6
合并请求
6
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
2d66aeeb
编写于
1月 18, 2023
作者:
G
gaotingquan
提交者:
Tingquan Gao
2月 14, 2023
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
support specify model_type in config
such as norm_train and to_static_train
上级
811b483e
变更
44
隐藏空白更改
内联
并排
Showing
44 changed file
with
235 addition
and
166 deletion
+235
-166
test_tipc/benchmark_train.sh
test_tipc/benchmark_train.sh
+136
-122
test_tipc/configs/CSWinTransformer/CSWinTransformer_tiny_224_train_infer_python.txt
...nsformer/CSWinTransformer_tiny_224_train_infer_python.txt
+1
-0
test_tipc/configs/DenseNet/DenseNet121_train_infer_python.txt
..._tipc/configs/DenseNet/DenseNet121_train_infer_python.txt
+7
-0
test_tipc/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_train_infer_python.txt
...on/GeneralRecognition_PPLCNet_x2_5_train_infer_python.txt
+2
-1
test_tipc/configs/GeneralRecognitionV2/GeneralRecognitionV2_PPLCNetV2_base_train_infer_python.txt
...eneralRecognitionV2_PPLCNetV2_base_train_infer_python.txt
+2
-1
test_tipc/configs/HRNet/HRNet_W18_C_train_infer_python.txt
test_tipc/configs/HRNet/HRNet_W18_C_train_infer_python.txt
+3
-2
test_tipc/configs/HRNet/HRNet_W48_C_train_infer_python.txt
test_tipc/configs/HRNet/HRNet_W48_C_train_infer_python.txt
+3
-2
test_tipc/configs/MobileNetV1/MobileNetV1_train_infer_python.txt
...pc/configs/MobileNetV1/MobileNetV1_train_infer_python.txt
+3
-2
test_tipc/configs/MobileNetV2/MobileNetV2_train_infer_python.txt
...pc/configs/MobileNetV2/MobileNetV2_train_infer_python.txt
+3
-2
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_infer_python.txt
...MobileNetV3/MobileNetV3_large_x1_0_train_infer_python.txt
+2
-1
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
...ain_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
+2
-1
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_pact_infer_python.txt
...eNetV3/MobileNetV3_large_x1_0_train_pact_infer_python.txt
+3
-2
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_ptq_infer_python.txt
...leNetV3/MobileNetV3_large_x1_0_train_ptq_infer_python.txt
+3
-2
test_tipc/configs/MobileViT/MobileViT_S_train_infer_python.txt
...tipc/configs/MobileViT/MobileViT_S_train_infer_python.txt
+1
-0
test_tipc/configs/PPHGNet/PPHGNet_small_train_infer_python.txt
...tipc/configs/PPHGNet/PPHGNet_small_train_infer_python.txt
+2
-1
test_tipc/configs/PPHGNet/PPHGNet_tiny_train_infer_python.txt
..._tipc/configs/PPHGNet/PPHGNet_tiny_train_infer_python.txt
+2
-1
test_tipc/configs/PPLCNet/PPLCNet_x0_25_train_infer_python.txt
...tipc/configs/PPLCNet/PPLCNet_x0_25_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x0_35_train_infer_python.txt
...tipc/configs/PPLCNet/PPLCNet_x0_35_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x0_5_train_infer_python.txt
..._tipc/configs/PPLCNet/PPLCNet_x0_5_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x0_75_train_infer_python.txt
...tipc/configs/PPLCNet/PPLCNet_x0_75_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x1_0_train_infer_python.txt
..._tipc/configs/PPLCNet/PPLCNet_x1_0_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x1_5_train_infer_python.txt
..._tipc/configs/PPLCNet/PPLCNet_x1_5_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x2_0_train_infer_python.txt
..._tipc/configs/PPLCNet/PPLCNet_x2_0_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNet/PPLCNet_x2_5_train_infer_python.txt
..._tipc/configs/PPLCNet/PPLCNet_x2_5_train_infer_python.txt
+1
-0
test_tipc/configs/PPLCNetV2/PPLCNetV2_base_train_infer_python.txt
...c/configs/PPLCNetV2/PPLCNetV2_base_train_infer_python.txt
+1
-0
test_tipc/configs/PVTV2/PVT_V2_B2_Linear_train_infer_python.txt
...ipc/configs/PVTV2/PVT_V2_B2_Linear_train_infer_python.txt
+1
-0
test_tipc/configs/ResNeXt/ResNeXt50_32x4d_train_infer_python.txt
...pc/configs/ResNeXt/ResNeXt50_32x4d_train_infer_python.txt
+7
-0
test_tipc/configs/ResNet/ResNet152_train_infer_python.txt
test_tipc/configs/ResNet/ResNet152_train_infer_python.txt
+3
-2
test_tipc/configs/ResNet/ResNet50_train_ampfp16_infer_python.txt
...pc/configs/ResNet/ResNet50_train_ampfp16_infer_python.txt
+2
-1
test_tipc/configs/ResNet/ResNet50_train_infer_python.txt
test_tipc/configs/ResNet/ResNet50_train_infer_python.txt
+3
-2
test_tipc/configs/ResNet/ResNet50_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
...ain_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
+2
-1
test_tipc/configs/ResNet/ResNet50_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
...train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
+1
-0
test_tipc/configs/ResNet/ResNet50_train_pact_infer_python.txt
..._tipc/configs/ResNet/ResNet50_train_pact_infer_python.txt
+2
-1
test_tipc/configs/ResNet/ResNet50_train_ptq_infer_python.txt
test_tipc/configs/ResNet/ResNet50_train_ptq_infer_python.txt
+3
-2
test_tipc/configs/ResNet/ResNet50_vd_train_infer_python.txt
test_tipc/configs/ResNet/ResNet50_vd_train_infer_python.txt
+2
-1
test_tipc/configs/ResNet/ResNet50_vd_train_pact_infer_python.txt
...pc/configs/ResNet/ResNet50_vd_train_pact_infer_python.txt
+2
-1
test_tipc/configs/ResNet/ResNet50_vd_train_ptq_infer_python.txt
...ipc/configs/ResNet/ResNet50_vd_train_ptq_infer_python.txt
+3
-2
test_tipc/configs/ShuffleNet/ShuffleNetV2_x1_0_train_infer_python.txt
...nfigs/ShuffleNet/ShuffleNetV2_x1_0_train_infer_python.txt
+3
-2
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_infer_python.txt
...ransformer_tiny_patch4_window7_224_train_infer_python.txt
+3
-2
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
...ain_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
+3
-2
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_pact_infer_python.txt
...ormer_tiny_patch4_window7_224_train_pact_infer_python.txt
+3
-2
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_ptq_infer_python.txt
...former_tiny_patch4_window7_224_train_ptq_infer_python.txt
+3
-2
test_tipc/configs/Twins/alt_gvt_base_train_infer_python.txt
test_tipc/configs/Twins/alt_gvt_base_train_infer_python.txt
+2
-1
test_tipc/configs/Twins/alt_gvt_small_train_infer_python.txt
test_tipc/configs/Twins/alt_gvt_small_train_infer_python.txt
+3
-2
未找到文件。
test_tipc/benchmark_train.sh
浏览文件 @
2d66aeeb
...
...
@@ -90,6 +90,8 @@ line_num=`expr $line_num + 1`
fp_items
=
$(
func_parser_value
"
${
lines
[line_num]
}
"
)
line_num
=
`
expr
$line_num
+ 1
`
epoch
=
$(
func_parser_value
"
${
lines
[line_num]
}
"
)
line_num
=
`
expr
$line_num
+ 1
`
model_type
=
$(
func_parser_value
"
${
lines
[line_num]
}
"
)
line_num
=
`
expr
$line_num
+ 1
`
profile_option_key
=
$(
func_parser_key
"
${
lines
[line_num]
}
"
)
...
...
@@ -118,6 +120,7 @@ line_gpuid=4
line_precision
=
6
line_epoch
=
7
line_batchsize
=
9
line_model_type
=
15
line_profile
=
13
line_eval_py
=
24
line_export_py
=
30
...
...
@@ -138,6 +141,7 @@ if [[ ! -n "$PARAMS" ]];then
batch_size_list
=(
${
batch_size
}
)
fp_items_list
=(
${
fp_items
}
)
device_num_list
=(
N1C4
)
model_type_list
=(
${
model_type
}
)
run_mode
=
"DP"
elif
[[
${
PARAMS
}
=
"dynamicTostatic"
]]
;
then
IFS
=
"|"
...
...
@@ -165,14 +169,13 @@ else
fp_items_list
=(
$precision
)
batch_size_list
=(
$batch_size
)
device_num_list
=(
$device_num
)
fi
# for log name
to_static
=
""
# parse "to_static" options and modify trainer into "to_static_trainer
"
if
[[
${
model_type
}
=
"dynamicTostatic"
]]
;
then
to_static
=
"d2sT_
"
sed
-i
's/trainer:norm_train/trainer:to_static_train/g'
$FILENAME
# parse "to_static" options and modify trainer into "to_static_trainer"
if
[[
${
model_type
}
=
"dynamicTostatic"
]]
;
then
model_type_list
=
"to_static_train
"
else
model_type_list
=
"norm_train
"
fi
fi
...
...
@@ -180,131 +183,142 @@ IFS="|"
for
batch_size
in
${
batch_size_list
[*]
}
;
do
for
precision
in
${
fp_items_list
[*]
}
;
do
for
device_num
in
${
device_num_list
[*]
}
;
do
# sed batchsize and precision
func_sed_params
"
$FILENAME
"
"
${
line_precision
}
"
"
$precision
"
func_sed_params
"
$FILENAME
"
"
${
line_batchsize
}
"
"
$batch_size
"
func_sed_params
"
$FILENAME
"
"
${
line_epoch
}
"
"
$epoch
"
gpu_id
=
$(
set_gpu_id
$device_num
)
for
model_type
in
${
model_type_list
[*]
}
;
do
# sed batchsize and precision
func_sed_params
"
$FILENAME
"
"
${
line_precision
}
"
"
$precision
"
func_sed_params
"
$FILENAME
"
"
${
line_batchsize
}
"
"
$batch_size
"
func_sed_params
"
$FILENAME
"
"
${
line_epoch
}
"
"
$epoch
"
func_sed_params
"
$FILENAME
"
"
${
line_model_type
}
"
"
$model_type
"
# It is needed that using dali, NHWC and 4 channels when training ResNet50 with AMPO2
if
[[
$model_name
==
"ResNet50"
&&
$precision
==
"fp16"
]]
;
then
sed
-i
"s/ResNet50.yaml/ResNet50_amp_O2_ultra.yaml/g"
$FILENAME
fi
# for log name
if
[[
${
model_type
}
=
"to_static_train"
]]
;
then
to_static
=
"d2sT_"
else
to_static
=
""
fi
# if bs is big, then copy train_list.txt to generate more train log
# At least 25 log number would be good to calculate ips for benchmark system.
# So the copy number for train_list is as follows:
total_batch_size
=
`
echo
$[$batch_size
*${
device_num
:1:1
}*${
device_num
:3:3
}
]
`
if
[[
$model_name
==
*
GeneralRecognition
*
]]
;
then
cd
dataset/
train_list_length
=
`
cat
train_reg_all_data.txt |
wc
-l
`
copy_num
=
`
echo
$[
25
*
10
*
$total_batch_size
/
$train_list_length
]
`
if
[[
$copy_num
-gt
1
]]
;
then
rm
-rf
train_reg_all_data.txt
for
((
i
=
1
;
i <
=
$copy_num
;
i++
))
;
do
cat
tipc_shitu_demo_data/demo_train.txt
>>
train_reg_all_data.txt
done
gpu_id
=
$(
set_gpu_id
$device_num
)
# It is needed that using dali, NHWC and 4 channels when training ResNet50 with AMPO2
if
[[
$model_name
==
"ResNet50"
&&
$precision
==
"fp16"
]]
;
then
sed
-i
"s/ResNet50.yaml/ResNet50_amp_O2_ultra.yaml/g"
$FILENAME
fi
cd
..
else
cd
dataset/ILSVRC2012
val_list_length
=
`
cat
val_list.txt |
wc
-l
`
copy_num
=
`
echo
$[
25
*
10
*
$total_batch_size
/
$val_list_length
]
`
rm
-rf
train_list.txt
if
[[
$copy_num
-gt
1
]]
;
then
for
((
i
=
1
;
i <
=
$copy_num
;
i++
))
;
do
cat
val_list.txt
>>
train_list.txt
done
# if bs is big, then copy train_list.txt to generate more train log
# At least 25 log number would be good to calculate ips for benchmark system.
# So the copy number for train_list is as follows:
total_batch_size
=
`
echo
$[$batch_size
*${
device_num
:1:1
}*${
device_num
:3:3
}
]
`
if
[[
$model_name
==
*
GeneralRecognition
*
]]
;
then
cd
dataset/
train_list_length
=
`
cat
train_reg_all_data.txt |
wc
-l
`
copy_num
=
`
echo
$[
25
*
10
*
$total_batch_size
/
$train_list_length
]
`
if
[[
$copy_num
-gt
1
]]
;
then
rm
-rf
train_reg_all_data.txt
for
((
i
=
1
;
i <
=
$copy_num
;
i++
))
;
do
cat
tipc_shitu_demo_data/demo_train.txt
>>
train_reg_all_data.txt
done
fi
cd
..
else
ln
-s
val_list.txt train_list.txt
cd
dataset/ILSVRC2012
val_list_length
=
`
cat
val_list.txt |
wc
-l
`
copy_num
=
`
echo
$[
25
*
10
*
$total_batch_size
/
$val_list_length
]
`
rm
-rf
train_list.txt
if
[[
$copy_num
-gt
1
]]
;
then
for
((
i
=
1
;
i <
=
$copy_num
;
i++
))
;
do
cat
val_list.txt
>>
train_list.txt
done
else
ln
-s
val_list.txt train_list.txt
fi
cd
../../
fi
cd
../../
fi
if
[[
${#
gpu_id
}
-le
1
]]
;
then
log_path
=
"
$SAVE_LOG
/profiling_log"
mkdir
-p
$log_path
log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
profiling"
func_sed_params
"
$FILENAME
"
"
${
line_gpuid
}
"
"0"
# sed used gpu_id
# set profile_option params
tmp
=
`
sed
-i
"
${
line_profile
}
s/.*/
${
profile_option
}
/"
"
${
FILENAME
}
"
`
if
[[
${#
gpu_id
}
-le
1
]]
;
then
log_path
=
"
$SAVE_LOG
/profiling_log"
mkdir
-p
$log_path
log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
profiling"
func_sed_params
"
$FILENAME
"
"
${
line_gpuid
}
"
"0"
# sed used gpu_id
# set profile_option params
tmp
=
`
sed
-i
"
${
line_profile
}
s/.*/
${
profile_option
}
/"
"
${
FILENAME
}
"
`
# run test_train_inference_python.sh
cmd
=
"bash test_tipc/test_train_inference_python.sh
${
FILENAME
}
benchmark_train >
${
log_path
}
/
${
log_name
}
2>&1 "
echo
$cmd
eval
$cmd
eval
"cat
${
log_path
}
/
${
log_name
}
"
# run test_train_inference_python.sh
cmd
=
"bash test_tipc/test_train_inference_python.sh
${
FILENAME
}
benchmark_train >
${
log_path
}
/
${
log_name
}
2>&1 "
echo
$cmd
eval
$cmd
eval
"cat
${
log_path
}
/
${
log_name
}
"
# without profile
log_path
=
"
$SAVE_LOG
/train_log"
speed_log_path
=
"
$SAVE_LOG
/index"
mkdir
-p
$log_path
mkdir
-p
$speed_log_path
log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
log"
speed_log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
speed"
func_sed_params
"
$FILENAME
"
"
${
line_profile
}
"
"null"
# sed profile_id as null
cmd
=
"bash test_tipc/test_train_inference_python.sh
${
FILENAME
}
benchmark_train >
${
log_path
}
/
${
log_name
}
2>&1 "
echo
$cmd
job_bt
=
`
date
'+%Y%m%d%H%M%S'
`
eval
$cmd
job_et
=
`
date
'+%Y%m%d%H%M%S'
`
export
model_run_time
=
$((${
job_et
}
-
${
job_bt
}))
eval
"cat
${
log_path
}
/
${
log_name
}
"
# without profile
log_path
=
"
$SAVE_LOG
/train_log"
speed_log_path
=
"
$SAVE_LOG
/index"
mkdir
-p
$log_path
mkdir
-p
$speed_log_path
log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
log"
speed_log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
speed"
func_sed_params
"
$FILENAME
"
"
${
line_profile
}
"
"null"
# sed profile_id as null
cmd
=
"bash test_tipc/test_train_inference_python.sh
${
FILENAME
}
benchmark_train >
${
log_path
}
/
${
log_name
}
2>&1 "
echo
$cmd
job_bt
=
`
date
'+%Y%m%d%H%M%S'
`
eval
$cmd
job_et
=
`
date
'+%Y%m%d%H%M%S'
`
export
model_run_time
=
$((${
job_et
}
-
${
job_bt
}))
eval
"cat
${
log_path
}
/
${
log_name
}
"
# parser log
_model_name
=
"
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
"
cmd
=
"
${
python
}
${
BENCHMARK_ROOT
}
/scripts/analysis.py --filename
${
log_path
}
/
${
log_name
}
\
--speed_log_file '
${
speed_log_path
}
/
${
speed_log_name
}
'
\
--model_name
${
_model_name
}
\
--base_batch_size
${
batch_size
}
\
--run_mode
${
run_mode
}
\
--fp_item
${
precision
}
\
--keyword ips:
\
--skip_steps 100
\
--device_num
${
device_num
}
\
--speed_unit samples/s
\
--convergence_key loss: "
echo
$cmd
eval
$cmd
last_status
=
${
PIPESTATUS
[0]
}
status_check
$last_status
"
${
cmd
}
"
"
${
status_log
}
"
"
${
model_name
}
"
else
IFS
=
";"
unset_env
=
`
unset
CUDA_VISIBLE_DEVICES
`
log_path
=
"
$SAVE_LOG
/train_log"
speed_log_path
=
"
$SAVE_LOG
/index"
mkdir
-p
$log_path
mkdir
-p
$speed_log_path
log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
log"
speed_log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
speed"
func_sed_params
"
$FILENAME
"
"
${
line_gpuid
}
"
"
$gpu_id
"
# sed used gpu_id
func_sed_params
"
$FILENAME
"
"
${
line_profile
}
"
"null"
# sed --profile_option as null
cmd
=
"bash test_tipc/test_train_inference_python.sh
${
FILENAME
}
benchmark_train >
${
log_path
}
/
${
log_name
}
2>&1 "
echo
$cmd
job_bt
=
`
date
'+%Y%m%d%H%M%S'
`
eval
$cmd
job_et
=
`
date
'+%Y%m%d%H%M%S'
`
export
model_run_time
=
$((${
job_et
}
-
${
job_bt
}))
eval
"cat
${
log_path
}
/
${
log_name
}
"
# parser log
_model_name
=
"
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
"
# parser log
_model_name
=
"
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
"
cmd
=
"
${
python
}
${
BENCHMARK_ROOT
}
/scripts/analysis.py --filename
${
log_path
}
/
${
log_name
}
\
--speed_log_file '
${
speed_log_path
}
/
${
speed_log_name
}
'
\
--model_name
${
_model_name
}
\
--base_batch_size
${
batch_size
}
\
--run_mode
${
run_mode
}
\
--fp_item
${
precision
}
\
--keyword ips:
\
--skip_steps 100
\
--device_num
${
device_num
}
\
--speed_unit samples/s
\
--convergence_key loss: "
echo
$cmd
eval
$cmd
last_status
=
${
PIPESTATUS
[0]
}
status_check
$last_status
"
${
cmd
}
"
"
${
status_log
}
"
"
${
model_name
}
"
else
IFS
=
";"
unset_env
=
`
unset
CUDA_VISIBLE_DEVICES
`
log_path
=
"
$SAVE_LOG
/train_log"
speed_log_path
=
"
$SAVE_LOG
/index"
mkdir
-p
$log_path
mkdir
-p
$speed_log_path
log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
log"
speed_log_name
=
"
${
repo_name
}
_
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
_
${
device_num
}
_
${
to_static
}
speed"
func_sed_params
"
$FILENAME
"
"
${
line_gpuid
}
"
"
$gpu_id
"
# sed used gpu_id
func_sed_params
"
$FILENAME
"
"
${
line_profile
}
"
"null"
# sed --profile_option as null
cmd
=
"bash test_tipc/test_train_inference_python.sh
${
FILENAME
}
benchmark_train >
${
log_path
}
/
${
log_name
}
2>&1 "
echo
$cmd
job_bt
=
`
date
'+%Y%m%d%H%M%S'
`
eval
$cmd
job_et
=
`
date
'+%Y%m%d%H%M%S'
`
export
model_run_time
=
$((${
job_et
}
-
${
job_bt
}))
eval
"cat
${
log_path
}
/
${
log_name
}
"
# parser log
_model_name
=
"
${
model_name
}
_bs
${
batch_size
}
_
${
precision
}
_
${
run_mode
}
"
cmd
=
"
${
python
}
${
BENCHMARK_ROOT
}
/scripts/analysis.py --filename
${
log_path
}
/
${
log_name
}
\
--speed_log_file '
${
speed_log_path
}
/
${
speed_log_name
}
'
\
--model_name
${
_model_name
}
\
--base_batch_size
${
batch_size
}
\
--run_mode
${
run_mode
}
\
--fp_item
${
precision
}
\
--keyword ips:
\
--skip_steps 100
\
--device_num
${
device_num
}
\
--speed_unit images/s
\
--convergence_key loss: "
echo
$cmd
eval
$cmd
last_status
=
${
PIPESTATUS
[0]
}
status_check
$last_status
"
${
cmd
}
"
"
${
status_log
}
"
"
${
model_name
}
"
fi
cmd
=
"
${
python
}
${
BENCHMARK_ROOT
}
/scripts/analysis.py --filename
${
log_path
}
/
${
log_name
}
\
--speed_log_file '
${
speed_log_path
}
/
${
speed_log_name
}
'
\
--model_name
${
_model_name
}
\
--base_batch_size
${
batch_size
}
\
--run_mode
${
run_mode
}
\
--fp_item
${
precision
}
\
--keyword ips:
\
--skip_steps 100
\
--device_num
${
device_num
}
\
--speed_unit images/s
\
--convergence_key loss: "
echo
$cmd
eval
$cmd
last_status
=
${
PIPESTATUS
[0]
}
status_check
$last_status
"
${
cmd
}
"
"
${
status_log
}
"
"
${
model_name
}
"
fi
done
done
done
done
test_tipc/configs/CSWinTransformer/CSWinTransformer_tiny_224_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,6 +54,7 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/DenseNet/DenseNet121_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -50,5 +50,12 @@ inference:python/predict_cls.py -c configs/inference_cls.yaml
-o Global.benchmark:False
null:null
null:null
===========================train_benchmark_params==========================
batch_size:64
fp_items:fp32|fp16
epoch:1
model_type:norm_train|to_static_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
test_tipc/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:256
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/GeneralRecognitionV2/GeneralRecognitionV2_PPLCNetV2_base_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/HRNet/HRNet_W18_C_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/HRNet/HRNet_W18_C.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/HRNet/HRNet_W48_C_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/HRNet/HRNet_W48_C.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64|128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileNetV1/MobileNetV1_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/MobileNetV1/MobileNetV1.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64|128
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileNetV2/MobileNetV2_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/MobileNetV2/MobileNetV2.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64|128
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:256|640
fp_items:fp32|fp16
epoch:1
model_type:norm_train|to_static_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:256|640
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_pact_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x1_0.yaml -o Slim.quant.name=pact
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:256|640
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_ptq_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/MobileNetV3/MobileNetV3_large_x1_0.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:256|640
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/MobileViT/MobileViT_S_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,6 +54,7 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPHGNet/PPHGNet_small_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/PPHGNet/PPHGNet_small.yaml
null:null
##
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:128
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPHGNet/PPHGNet_tiny_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/PPHGNet/PPHGNet_tiny.yaml
null:null
##
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:128
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x0_25_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x0_35_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x0_5_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x0_75_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x1_0_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x1_5_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x2_0_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:512
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNet/PPLCNet_x2_5_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:256
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PPLCNetV2/PPLCNetV2_base_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -53,6 +53,7 @@ null:null
batch_size:500
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/PVTV2/PVT_V2_B2_Linear_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,6 +54,7 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/ResNeXt/ResNeXt50_32x4d_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -50,5 +50,12 @@ inference:python/predict_cls.py -c configs/inference_cls.yaml
-o Global.benchmark:False
null:null
null:null
===========================train_benchmark_params==========================
batch_size:64
fp_items:fp32|fp16
epoch:1
model_type:norm_train|to_static_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
test_tipc/configs/ResNet/ResNet152_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/ResNet/ResNet152.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:32|64
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_train_ampfp16_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/ResNet/ResNet50.yaml
null:null
##
...
...
@@ -54,3 +54,4 @@ null:null
batch_size:128|256
fp_items:ampfp16
epoch:1
model_type:norm_train
test_tipc/configs/ResNet/ResNet50_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/ResNet/ResNet50.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128|64
fp_items:fp32|fp16
epoch:1
model_type:norm_train|to_static_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,3 +54,4 @@ null:null
batch_size:128|256
fp_items:purefp16
epoch:1
model_type:norm_train
test_tipc/configs/ResNet/ResNet50_train_pact_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_train_ptq_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/ResNet/ResNet50.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_vd_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_vd_train_pact_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ResNet/ResNet50_vd_train_ptq_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/ResNet/ResNet50_vd.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/ShuffleNet/ShuffleNetV2_x1_0_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
to_static_train:-o Global.to_static=True
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/ShuffleNet/ShuffleNetV2_x1_0.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:256|1536
fp_items:fp32|fp16
epoch:2
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/SwinTransformer/SwinTransformer_tiny_patch4_window7_224.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:104|128
fp_items:fp32|fp16
epoch:1
model_type:norm_train|to_static_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/SwinTransformer/SwinTransformer_tiny_patch4_window7_224.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64|104
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_pact_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/SwinTransformer/SwinTransformer_tiny_patch4_window7_224.yaml -o Slim.quant.name=pact
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64|104
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_ptq_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/SwinTransformer/SwinTransformer_tiny_patch4_window7_224.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:64|104
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
test_tipc/configs/Twins/alt_gvt_base_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/Twins/alt_gvt_base.yaml
null:null
##
...
...
@@ -54,6 +54,7 @@ null:null
batch_size:64|144
fp_items:fp32
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
...
...
test_tipc/configs/Twins/alt_gvt_small_train_infer_python.txt
浏览文件 @
2d66aeeb
...
...
@@ -20,7 +20,7 @@ distill_train:null
null:null
null:null
##
===========================eval_params===========================
===========================eval_params===========================
eval:tools/eval.py -c ppcls/configs/ImageNet/Twins/alt_gvt_small.yaml
null:null
##
...
...
@@ -54,7 +54,8 @@ null:null
batch_size:128
fp_items:fp32|fp16
epoch:1
model_type:norm_train
--profiler_options:batch_range=[10,20];state=GPU;tracer_option=Default;profile_path=model.profile
flags:FLAGS_eager_delete_tensor_gb=0.0;FLAGS_fraction_of_gpu_memory_to_use=0.98;FLAGS_conv_workspace_size_limit=4096
===========================infer_benchmark_params==========================
random_infer_input:[{float32,[3,224,224]}]
\ No newline at end of file
random_infer_input:[{float32,[3,224,224]}]
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