test to test_v5
This commit is contained in:
LDOUBLEV
parent
3ba4d543a6
commit
069d994c16
173
test/infer.sh
173
test/infer.sh
|
|
@ -1,173 +0,0 @@
|
|||
#!/bin/bash
|
||||
FILENAME=$1
|
||||
dataline=$(cat ${FILENAME})
|
||||
# parser params
|
||||
IFS=$'\n'
|
||||
lines=(${dataline})
|
||||
function func_parser(){
|
||||
strs=$1
|
||||
IFS=": "
|
||||
array=(${strs})
|
||||
tmp=${array[1]}
|
||||
echo ${tmp}
|
||||
}
|
||||
IFS=$'\n'
|
||||
# The training params
|
||||
train_model_list=$(func_parser "${lines[0]}")
|
||||
slim_trainer_list=$(func_parser "${lines[3]}")
|
||||
python=$(func_parser "${lines[4]}")
|
||||
# inference params
|
||||
# inference=$(func_parser "${lines[5]}")
|
||||
devices=$(func_parser "${lines[6]}")
|
||||
use_mkldnn_list=$(func_parser "${lines[7]}")
|
||||
cpu_threads_list=$(func_parser "${lines[8]}")
|
||||
rec_batch_size_list=$(func_parser "${lines[9]}")
|
||||
gpu_trt_list=$(func_parser "${lines[10]}")
|
||||
gpu_precision_list=$(func_parser "${lines[11]}")
|
||||
|
||||
infer_gpu_id=$(func_parser "${lines[12]}")
|
||||
log_path=$(func_parser "${lines[13]}")
|
||||
status_log="${log_path}/result.log"
|
||||
|
||||
# install requirments
|
||||
${python} -m pip install pynvml;
|
||||
${python} -m pip install psutil;
|
||||
${python} -m pip install GPUtil;
|
||||
|
||||
paddle_info="$(${python} -c "import paddle;print(f'paddle_version:{paddle.__version__}');print(f'paddle_commit:{paddle.__git_commit__}')")"
|
||||
echo -e "\033[33m $paddle_info \033[0m" | tee -a ${status_log}
|
||||
cpu_model=`cat /proc/cpuinfo | grep "model name" | awk -F ':' '{print $2}' | sort | uniq`
|
||||
echo -e "\033[33m cpu_info:$cpu_model \033[0m" | tee -a ${status_log}
|
||||
ip=`ifconfig| grep -A 1 'eth0'|grep 'inet'|awk -F ':' '{print $2}'|awk '{print $1}'`
|
||||
echo -e "\033[33m ip_info:$ip \033[0m" | tee -a ${status_log}
|
||||
|
||||
function status_check(){
|
||||
last_status=$1 # the exit code
|
||||
run_model=$2
|
||||
run_command=$3
|
||||
run_log=$4
|
||||
if [ $last_status -eq 0 ]; then
|
||||
echo -e "\033[33m $run_model successfully with command - ${run_command}! \033[0m" | tee -a ${run_log}
|
||||
else
|
||||
echo -e "\033[33m $case failed with command - ${run_command}! \033[0m" | tee -a ${run_log}
|
||||
fi
|
||||
}
|
||||
IFS='|'
|
||||
for train_model in ${train_model_list[*]}; do
|
||||
if [ ${train_model} = "ocr_det" ];then
|
||||
model_name="det"
|
||||
yml_file="configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar
|
||||
cd ./inference && tar xf ch_det_data_50.tar && cd ../
|
||||
img_dir="./inference/ch_det_data_50/all-sum-510"
|
||||
data_dir=./inference/ch_det_data_50/
|
||||
data_label_file=[./inference/ch_det_data_50/test_gt_50.txt]
|
||||
elif [ ${train_model} = "ocr_rec" ];then
|
||||
model_name="rec"
|
||||
yml_file="configs/rec/rec_mv3_none_bilstm_ctc.yml"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_rec_data_200.tar
|
||||
cd ./inference && tar xf ch_rec_data_200.tar && cd ../
|
||||
img_dir="./inference/ch_rec_data_200/"
|
||||
fi
|
||||
|
||||
# eval
|
||||
for slim_trainer in ${slim_trainer_list[*]}; do
|
||||
if [ ${slim_trainer} = "norm" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
elif [ ${slim_trainer} = "pact" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_quant_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_quant_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_quant_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_rec_quant_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
elif [ ${slim_trainer} = "distill" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_distill_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_distill_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_distill_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_rec_distill_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
elif [ ${slim_trainer} = "fpgm" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_prune_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_prune_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_prune_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_rec_prune_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
fi
|
||||
|
||||
save_log_path="${log_path}/${eval_model_name}"
|
||||
command="${python} tools/eval.py -c ${yml_file} -o Global.pretrained_model='./inference/${eval_model_name}/best_accuracy' Global.save_model_dir=${save_log_path} Eval.dataset.data_dir=${data_dir} Eval.dataset.label_file_list=${data_label_file}"
|
||||
${python} tools/eval.py -c ${yml_file} -o Global.pretrained_model=./inference/${eval_model_name}/best_accuracy Global.save_model_dir=${save_log_path} Eval.dataset.data_dir=${data_dir} Eval.dataset.label_file_list=${data_label_file}
|
||||
status_check $? "${trainer}" "${command}" "${status_log}"
|
||||
|
||||
command="${python} tools/export_model.py -c ${yml_file} -o Global.pretrained_model="${eval_model_name}/best_accuracy" Global.save_inference_dir=${log_path}/${eval_model_name}_infer Global.save_model_dir=${save_log_path}"
|
||||
${python} tools/export_model.py -c ${yml_file} -o Global.pretrained_model="./inference/${eval_model_name}/best_accuracy" Global.save_inference_dir="${log_path}/${eval_model_name}_infer" Global.save_model_dir=${save_log_path}
|
||||
status_check $? "${trainer}" "${command}" "${status_log}"
|
||||
|
||||
if [ $? -eq 0 ]; then
|
||||
echo -e "\033[33m training of $model_name successfully!\033[0m" | tee -a ${save_log}/train.log
|
||||
else
|
||||
cat ${save_log}/train.log
|
||||
echo -e "\033[33m training of $model_name failed!\033[0m" | tee -a ${save_log}/train.log
|
||||
fi
|
||||
if [ "${model_name}" = "det" ]; then
|
||||
export rec_batch_size_list=( "1" )
|
||||
inference="tools/infer/predict_det.py"
|
||||
det_model_dir="${log_path}/${eval_model_name}_infer"
|
||||
rec_model_dir=""
|
||||
elif [ "${model_name}" = "rec" ]; then
|
||||
inference="tools/infer/predict_rec.py"
|
||||
rec_model_dir="${log_path}/${eval_model_name}_infer"
|
||||
det_model_dir=""
|
||||
fi
|
||||
# inference
|
||||
for device in ${devices[*]}; do
|
||||
if [ ${device} = "cpu" ]; then
|
||||
for use_mkldnn in ${use_mkldnn_list[*]}; do
|
||||
for threads in ${cpu_threads_list[*]}; do
|
||||
for rec_batch_size in ${rec_batch_size_list[*]}; do
|
||||
save_log_path="${log_path}/${model_name}_${slim_trainer}_cpu_usemkldnn_${use_mkldnn}_cputhreads_${threads}_recbatchnum_${rec_batch_size}_infer.log"
|
||||
command="${python} ${inference} --enable_mkldnn=${use_mkldnn} --use_gpu=False --cpu_threads=${threads} --benchmark=True --det_model_dir=${det_model_dir} --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}"
|
||||
${python} ${inference} --enable_mkldnn=${use_mkldnn} --use_gpu=False --cpu_threads=${threads} --benchmark=True --det_model_dir=${det_model_dir} --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}
|
||||
status_check $? "${trainer}" "${command}" "${status_log}"
|
||||
done
|
||||
done
|
||||
done
|
||||
else
|
||||
# env="export CUDA_VISIBLE_DEVICES=${infer_gpu_id}"
|
||||
for use_trt in ${gpu_trt_list[*]}; do
|
||||
for precision in ${gpu_precision_list[*]}; do
|
||||
if [ ${use_trt} = "False" ] && [ ${precision} != "fp32" ]; then
|
||||
continue
|
||||
fi
|
||||
for rec_batch_size in ${rec_batch_size_list[*]}; do
|
||||
save_log_path="${log_path}/${model_name}_${slim_trainer}_gpu_usetensorrt_${use_trt}_usefp16_${precision}_recbatchnum_${rec_batch_size}_infer.log"
|
||||
command="${python} ${inference} --use_gpu=True --use_tensorrt=${use_trt} --precision=${precision} --benchmark=True --det_model_dir=${log_path}/${eval_model_name}_infer --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}"
|
||||
${python} ${inference} --use_gpu=True --use_tensorrt=${use_trt} --precision=${precision} --benchmark=True --det_model_dir=${log_path}/${eval_model_name}_infer --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}
|
||||
status_check $? "${trainer}" "${command}" "${status_log}"
|
||||
done
|
||||
done
|
||||
done
|
||||
fi
|
||||
done
|
||||
done
|
||||
done
|
||||
|
|
@ -0,0 +1,35 @@
|
|||
model_name:ocr_det
|
||||
python:python3.7
|
||||
gpu_list:-1|0|0,1
|
||||
Global.auto_cast:False|True
|
||||
Global.epoch_num:10
|
||||
Global.save_model_dir:./output/
|
||||
Global.save_inference_dir:./output/
|
||||
Train.loader.batch_size_per_card:
|
||||
Global.use_gpu
|
||||
Global.pretrained_model
|
||||
|
||||
trainer:norm|pact|fpgm
|
||||
norm_train:tools/train.py -c configs/det/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/MobileNetV3_large_x0_5_pretrained
|
||||
quant_train:deploy/slim/quantization/quant.py -c configs/det/det_mv3_db.yml -o Global.pretrained_model=./pretrain_models/det_mv3_db_v2.0_train/best_accuracy
|
||||
fpgm_train:null
|
||||
distill_train:null
|
||||
|
||||
eval:tools/eval.py -c configs/det/det_mv3_db.yml -o
|
||||
|
||||
norm_export:tools/export_model.py -c configs/det/det_mv3_db.yml -o
|
||||
quant_export:deploy/slim/quantization/export_model.py -c configs/det/det_mv3_db.yml -o
|
||||
fpgm_export:deploy/slim/prune/export_prune_model.py
|
||||
distill_export:null
|
||||
|
||||
inference:tools/infer/predict_det.py
|
||||
--use_gpu:True|False
|
||||
--enable_mkldnn:True|False
|
||||
--cpu_threads:1|6
|
||||
--rec_batch_num:1
|
||||
--use_tensorrt:True|False
|
||||
--precision:fp32|fp16|int8
|
||||
--det_model_dir
|
||||
--image_dir
|
||||
--save_log_path
|
||||
|
||||
|
|
@ -1,15 +0,0 @@
|
|||
train_model_list: ocr_det
|
||||
gpu_list: -1|0|0,1
|
||||
auto_cast_list: False|True
|
||||
trainer_list: norm|pact|fpgm
|
||||
python: python3.7
|
||||
|
||||
inference: python
|
||||
devices: cpu|gpu
|
||||
use_mkldnn_list: True|False
|
||||
cpu_threads_list: 1|6
|
||||
rec_batch_size_list: 1|6
|
||||
gpu_trt_list: True|False
|
||||
gpu_precision_list: fp32|fp16|int8
|
||||
infer_gpu_id: 0
|
||||
log_path: ./output
|
||||
|
|
@ -0,0 +1,138 @@
|
|||
#!/bin/bash
|
||||
FILENAME=$1
|
||||
# MODE be one of ['lite_train_infer' 'whole_infer' 'whole_train_infer', 'infer']
|
||||
MODE=$2
|
||||
|
||||
dataline=$(cat ${FILENAME})
|
||||
|
||||
# parser params
|
||||
IFS=$'\n'
|
||||
lines=(${dataline})
|
||||
function func_parser_key(){
|
||||
strs=$1
|
||||
IFS=":"
|
||||
array=(${strs})
|
||||
tmp=${array[0]}
|
||||
echo ${tmp}
|
||||
}
|
||||
function func_parser_value(){
|
||||
strs=$1
|
||||
IFS=":"
|
||||
array=(${strs})
|
||||
tmp=${array[1]}
|
||||
echo ${tmp}
|
||||
}
|
||||
IFS=$'\n'
|
||||
# The training params
|
||||
model_name=$(func_parser_value "${lines[0]}")
|
||||
train_model_list=$(func_parser_value "${lines[0]}")
|
||||
slim_trainer_list=$(func_parser_value "${lines[12]}")
|
||||
|
||||
# MODE be one of ['lite_train_infer' 'whole_infer' 'whole_train_infer']
|
||||
MODE=$2
|
||||
# prepare pretrained weights and dataset
|
||||
wget -nc -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams
|
||||
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar
|
||||
cd pretrain_models && tar xf det_mv3_db_v2.0_train.tar && cd ../
|
||||
|
||||
if [ ${MODE} = "lite_train_infer" ];then
|
||||
# pretrain lite train data
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_lite.tar
|
||||
cd ./train_data/ && tar xf icdar2015_lite.tar
|
||||
ln -s ./icdar2015_lite ./icdar2015
|
||||
cd ../
|
||||
epoch=10
|
||||
eval_batch_step=10
|
||||
elif [ ${MODE} = "whole_train_infer" ];then
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015.tar
|
||||
cd ./train_data/ && tar xf icdar2015.tar && cd ../
|
||||
epoch=500
|
||||
eval_batch_step=200
|
||||
elif [ ${MODE} = "whole_infer" ];then
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_infer.tar
|
||||
cd ./train_data/ && tar xf icdar2015_infer.tar
|
||||
ln -s ./icdar2015_infer ./icdar2015
|
||||
cd ../
|
||||
epoch=10
|
||||
eval_batch_step=10
|
||||
else
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar
|
||||
if [ ${model_name} = "ocr_det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
fi
|
||||
|
||||
|
||||
IFS='|'
|
||||
for train_model in ${train_model_list[*]}; do
|
||||
if [ ${train_model} = "ocr_det" ];then
|
||||
model_name="det"
|
||||
yml_file="configs/det/ch_ppocr_v2.0/ch_det_mv3_db_v2.0.yml"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_det_data_50.tar
|
||||
cd ./inference && tar xf ch_det_data_50.tar && cd ../
|
||||
img_dir="./inference/ch_det_data_50/all-sum-510"
|
||||
data_dir=./inference/ch_det_data_50/
|
||||
data_label_file=[./inference/ch_det_data_50/test_gt_50.txt]
|
||||
elif [ ${train_model} = "ocr_rec" ];then
|
||||
model_name="rec"
|
||||
yml_file="configs/rec/rec_mv3_none_bilstm_ctc.yml"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/ch_rec_data_200.tar
|
||||
cd ./inference && tar xf ch_rec_data_200.tar && cd ../
|
||||
img_dir="./inference/ch_rec_data_200/"
|
||||
fi
|
||||
|
||||
# eval
|
||||
for slim_trainer in ${slim_trainer_list[*]}; do
|
||||
if [ ${slim_trainer} = "norm" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_det_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/ch/ch_ppocr_mobile_v2.0_rec_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
elif [ ${slim_trainer} = "pact" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_quant_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_quant_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_quant_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_rec_quant_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
elif [ ${slim_trainer} = "distill" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_distill_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_distill_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_distill_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_rec_distill_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
elif [ ${slim_trainer} = "fpgm" ]; then
|
||||
if [ ${model_name} = "det" ]; then
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_det_prune_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_det_prune_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
else
|
||||
eval_model_name="ch_ppocr_mobile_v2.0_rec_prune_train"
|
||||
wget -nc -P ./inference https://paddleocr.bj.bcebos.com/dygraph_v2.0/slim/ch_ppocr_mobile_v2.0_rec_prune_train.tar
|
||||
cd ./inference && tar xf ${eval_model_name}.tar && cd ../
|
||||
fi
|
||||
fi
|
||||
done
|
||||
done
|
||||
374
test/test.sh
374
test/test.sh
|
|
@ -1,203 +1,221 @@
|
|||
#!/bin/bash
|
||||
# Usage:
|
||||
# bash test/test.sh ./test/paddleocr_ci_params.txt 'lite_train_infer'
|
||||
|
||||
#!/bin/bash
|
||||
FILENAME=$1
|
||||
|
||||
# MODE be one of ['lite_train_infer' 'whole_infer' 'whole_train_infer']
|
||||
# MODE be one of ['lite_train_infer' 'whole_infer' 'whole_train_infer', 'infer']
|
||||
MODE=$2
|
||||
# prepare pretrained weights and dataset
|
||||
wget -nc -P ./pretrain_models/ https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/MobileNetV3_large_x0_5_pretrained.pdparams
|
||||
wget -nc -P ./pretrain_models/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/en/det_mv3_db_v2.0_train.tar
|
||||
cd pretrain_models && tar xf det_mv3_db_v2.0_train.tar && cd ../
|
||||
|
||||
if [ ${MODE} = "lite_train_infer" ];then
|
||||
# pretrain lite train data
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_lite.tar
|
||||
cd ./train_data/ && tar xf icdar2015_lite.tar
|
||||
ln -s ./icdar2015_lite ./icdar2015
|
||||
cd ../
|
||||
epoch=10
|
||||
eval_batch_step=10
|
||||
elif [ ${MODE} = "whole_train_infer" ];then
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015.tar
|
||||
cd ./train_data/ && tar xf icdar2015.tar && cd ../
|
||||
epoch=500
|
||||
eval_batch_step=200
|
||||
else
|
||||
rm -rf ./train_data/icdar2015
|
||||
wget -nc -P ./train_data/ https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/icdar2015_infer.tar
|
||||
cd ./train_data/ && tar xf icdar2015_infer.tar
|
||||
ln -s ./icdar2015_infer ./icdar2015
|
||||
cd ../
|
||||
epoch=10
|
||||
eval_batch_step=10
|
||||
fi
|
||||
|
||||
img_dir="./train_data/icdar2015/text_localization/ch4_test_images/"
|
||||
|
||||
|
||||
dataline=$(cat ${FILENAME})
|
||||
|
||||
# parser params
|
||||
IFS=$'\n'
|
||||
lines=(${dataline})
|
||||
function func_parser(){
|
||||
function func_parser_key(){
|
||||
strs=$1
|
||||
IFS=": "
|
||||
IFS=":"
|
||||
array=(${strs})
|
||||
tmp=${array[0]}
|
||||
echo ${tmp}
|
||||
}
|
||||
function func_parser_value(){
|
||||
strs=$1
|
||||
IFS=":"
|
||||
array=(${strs})
|
||||
tmp=${array[1]}
|
||||
echo ${tmp}
|
||||
}
|
||||
IFS=$'\n'
|
||||
# The training params
|
||||
train_model_list=$(func_parser "${lines[0]}")
|
||||
gpu_list=$(func_parser "${lines[1]}")
|
||||
auto_cast_list=$(func_parser "${lines[2]}")
|
||||
slim_trainer_list=$(func_parser "${lines[3]}")
|
||||
python=$(func_parser "${lines[4]}")
|
||||
# inference params
|
||||
inference=$(func_parser "${lines[5]}")
|
||||
devices=$(func_parser "${lines[6]}")
|
||||
use_mkldnn_list=$(func_parser "${lines[7]}")
|
||||
cpu_threads_list=$(func_parser "${lines[8]}")
|
||||
rec_batch_size_list=$(func_parser "${lines[9]}")
|
||||
gpu_trt_list=$(func_parser "${lines[10]}")
|
||||
gpu_precision_list=$(func_parser "${lines[11]}")
|
||||
|
||||
log_path=$(func_parser "${lines[13]}")
|
||||
status_log="${log_path}/result.log"
|
||||
|
||||
# install requirments
|
||||
${python} -m pip install pynvml;
|
||||
${python} -m pip install psutil;
|
||||
${python} -m pip install GPUtil;
|
||||
${python} -m pip install paddlesim==2.0.0
|
||||
|
||||
paddle_info="$(${python} -c "import paddle;print(f'paddle_version:{paddle.__version__}');print(f'paddle_commit:{paddle.__git_commit__}')")"
|
||||
echo -e "\033[33m $paddle_info \033[0m" | tee -a ${status_log}
|
||||
cpu_model=`cat /proc/cpuinfo | grep "model name" | awk -F ':' '{print $2}' | sort | uniq`
|
||||
echo -e "\033[33m cpu_info:$cpu_model \033[0m" | tee -a ${status_log}
|
||||
ip=`ifconfig| grep -A 1 'eth0'|grep 'inet'|awk -F ':' '{print $2}'|awk '{print $1}'`
|
||||
echo -e "\033[33m ip_info:$ip \033[0m" | tee -a ${status_log}
|
||||
|
||||
function status_check(){
|
||||
last_status=$1 # the exit code
|
||||
run_model=$2
|
||||
run_command=$3
|
||||
run_log=$4
|
||||
run_command=$2
|
||||
run_log=$3
|
||||
if [ $last_status -eq 0 ]; then
|
||||
echo -e "\033[33m $run_model successfully with command - ${run_command}! \033[0m" | tee -a ${run_log}
|
||||
echo -e "\033[33m Run successfully with command - ${run_command}! \033[0m" | tee -a ${run_log}
|
||||
else
|
||||
echo -e "\033[33m $case failed with command - ${run_command}! \033[0m" | tee -a ${run_log}
|
||||
echo -e "\033[33m Run failed with command - ${run_command}! \033[0m" | tee -a ${run_log}
|
||||
fi
|
||||
}
|
||||
|
||||
IFS="|"
|
||||
for train_model in ${train_model_list[*]}; do
|
||||
if [ ${train_model} = "ocr_det" ];then
|
||||
model_name="det"
|
||||
yml_file="configs/det/det_mv3_db.yml"
|
||||
elif [ ${train_model} = "ocr_rec" ];then
|
||||
model_name="rec"
|
||||
yml_file="configs/rec/rec_mv3_none_bilstm_ctc.yml"
|
||||
else
|
||||
model_name="det"
|
||||
yml_file="configs/det/det_mv3_db.yml"
|
||||
fi
|
||||
IFS="|"
|
||||
for gpu in ${gpu_list[*]}; do
|
||||
use_gpu=True
|
||||
if [ ${gpu} = "-1" ];then
|
||||
use_gpu=False
|
||||
env=""
|
||||
elif [ ${#gpu} -le 1 ];then
|
||||
env="CUDA_VISIBLE_DEVICES=${gpu}"
|
||||
else
|
||||
IFS=","
|
||||
array=(${gpu})
|
||||
env="CUDA_VISIBLE_DEVICES=${array[0]}"
|
||||
IFS="|"
|
||||
fi
|
||||
for auto_cast in ${auto_cast_list[*]}; do
|
||||
for slim_trainer in ${slim_trainer_list[*]}; do
|
||||
if [ ${slim_trainer} = "norm" ]; then
|
||||
trainer="tools/train.py"
|
||||
export_model="tools/export_model.py"
|
||||
pretrain="./pretrain_models/MobileNetV3_large_x0_5_pretrained"
|
||||
elif [ ${slim_trainer} = "pact" ]; then
|
||||
trainer="deploy/slim/quantization/quant.py"
|
||||
export_model="deploy/slim/quantization/export_model.py"
|
||||
pretrain="./pretrain_models/det_mv3_db_v2.0_train/best_accuracy"
|
||||
elif [ ${slim_trainer} = "fpgm" ]; then
|
||||
trainer="deploy/slim/prune/sensitivity_anal.py"
|
||||
export_model="deploy/slim/prune/export_prune_model.py"
|
||||
pretrain="./pretrain_models/det_mv3_db_v2.0_train/best_accuracy"
|
||||
wget -nc -P https://paddleocr.bj.bcebos.com/dygraph_v2.0/test/sen.pickle
|
||||
elif [ ${slim_trainer} = "distill" ]; then
|
||||
trainer="deploy/slim/distill/train_dml.py"
|
||||
export_model="deploy/slim/distill/export_distill_model.py"
|
||||
pretrain=""
|
||||
else
|
||||
trainer="tools/train.py"
|
||||
export_model="tools/export_model.py"
|
||||
pretrain="./pretrain_models/MobileNetV3_large_x0_5_pretrained"
|
||||
fi
|
||||
save_log="${log_path}/${model_name}_${slim_trainer}_autocast_${auto_cast}_gpuid_${gpu}"
|
||||
if [ ${#gpu} -le 2 ];then
|
||||
command="${python} ${trainer} -c ${yml_file} -o Global.epoch_num=${epoch} Global.eval_batch_step=${eval_batch_step} Global.auto_cast=${auto_cast} Global.pretrained_model=${pretrain} Global.save_model_dir=${save_log} Global.use_gpu=${use_gpu} Train.loader.batch_size_per_card=2"
|
||||
${python} ${trainer} -c ${yml_file} -o Global.epoch_num=${epoch} Global.eval_batch_step=${eval_batch_step} Global.auto_cast=${auto_cast} Global.pretrained_model=${pretrain} Global.save_model_dir=${save_log} Global.use_gpu=${use_gpu} Train.loader.batch_size_per_card=2
|
||||
else
|
||||
command="${python} -m paddle.distributed.launch --log_dir=./debug/ --gpus ${gpu} ${trainer} -c ${yml_file} -o Global.epoch_num=${epoch} Global.eval_batch_step=${eval_batch_step} Global.auto_cast=${auto_cast} Global.pretrained_model=${pretrain} Global.save_model_dir=${save_log} Global.use_gpu=${use_gpu} Train.loader.batch_size_per_card=2"
|
||||
${python} -m paddle.distributed.launch --log_dir=./debug/ --gpus ${gpu} ${trainer} -c ${yml_file} -o Global.epoch_num=${epoch} Global.eval_batch_step=${eval_batch_step} Global.auto_cast=${auto_cast} Global.pretrained_model=${pretrain} Global.save_model_dir=${save_log} Global.use_gpu=${use_gpu} Train.loader.batch_size_per_card=2
|
||||
fi
|
||||
status_check $? "${trainer}" "${command}" "${status_log}"
|
||||
IFS=$'\n'
|
||||
# The training params
|
||||
model_name=$(func_parser_value "${lines[0]}")
|
||||
python=$(func_parser_value "${lines[1]}")
|
||||
gpu_list=$(func_parser_value "${lines[2]}")
|
||||
autocast_list=$(func_parser_value "${lines[3]}")
|
||||
autocast_key=$(func_parser_key "${lines[3]}")
|
||||
epoch_key=$(func_parser_key "${lines[4]}")
|
||||
save_model_key=$(func_parser_key "${lines[5]}")
|
||||
save_infer_key=$(func_parser_key "${lines[6]}")
|
||||
train_batch_key=$(func_parser_key "${lines[7]}")
|
||||
train_use_gpu_key=$(func_parser_key "${lines[8]}")
|
||||
pretrain_model_key=$(func_parser_key "${lines[9]}")
|
||||
|
||||
command="${python} ${export_model} -c ${yml_file} -o Global.pretrained_model=${save_log}/latest Global.save_inference_dir=${save_log}_infer/ Global.save_model_dir=${save_log}"
|
||||
${python} ${export_model} -c ${yml_file} -o Global.pretrained_model=${save_log}/latest Global.save_inference_dir=${save_log}_infer/ Global.save_model_dir=${save_log}
|
||||
status_check $? "${trainer}" "${command}" "${status_log}"
|
||||
|
||||
if [ "${model_name}" = "det" ]; then
|
||||
export rec_batch_size_list=( "1" )
|
||||
inference="tools/infer/predict_det.py"
|
||||
det_model_dir=${save_log}_infer
|
||||
rec_model_dir=""
|
||||
elif [ "${model_name}" = "rec" ]; then
|
||||
inference="tools/infer/predict_rec.py"
|
||||
rec_model_dir=${save_log}_infer
|
||||
det_model_dir=""
|
||||
fi
|
||||
# inference
|
||||
for device in ${devices[*]}; do
|
||||
if [ ${device} = "cpu" ]; then
|
||||
for use_mkldnn in ${use_mkldnn_list[*]}; do
|
||||
for threads in ${cpu_threads_list[*]}; do
|
||||
for rec_batch_size in ${rec_batch_size_list[*]}; do
|
||||
save_log_path="${log_path}/${model_name}_${slim_trainer}_cpu_usemkldnn_${use_mkldnn}_cputhreads_${threads}_recbatchnum_${rec_batch_size}_infer.log"
|
||||
command="${python} ${inference} --enable_mkldnn=${use_mkldnn} --use_gpu=False --cpu_threads=${threads} --benchmark=True --det_model_dir=${det_model_dir} --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}"
|
||||
${python} ${inference} --enable_mkldnn=${use_mkldnn} --use_gpu=False --cpu_threads=${threads} --benchmark=True --det_model_dir=${det_model_dir} --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}
|
||||
status_check $? "${inference}" "${command}" "${status_log}"
|
||||
done
|
||||
done
|
||||
done
|
||||
else
|
||||
for use_trt in ${gpu_trt_list[*]}; do
|
||||
for precision in ${gpu_precision_list[*]}; do
|
||||
if [ ${use_trt} = "False" ] && [ ${precision} != "fp32" ]; then
|
||||
continue
|
||||
fi
|
||||
for rec_batch_size in ${rec_batch_size_list[*]}; do
|
||||
save_log_path="${log_path}/${model_name}_${slim_trainer}_gpu_usetensorrt_${use_trt}_usefp16_${precision}_recbatchnum_${rec_batch_size}_infer.log"
|
||||
command="${python} ${inference} --use_gpu=True --use_tensorrt=${use_trt} --precision=${precision} --benchmark=True --det_model_dir=${det_model_dir} --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}"
|
||||
${python} ${inference} --use_gpu=True --use_tensorrt=${use_trt} --precision=${precision} --benchmark=True --det_model_dir=${det_model_dir} --rec_batch_num=${rec_batch_size} --rec_model_dir=${rec_model_dir} --image_dir=${img_dir} --save_log_path=${save_log_path}
|
||||
status_check $? "${inference}" "${command}" "${status_log}"
|
||||
done
|
||||
done
|
||||
done
|
||||
fi
|
||||
trainer_list=$(func_parser_value "${lines[10]}")
|
||||
norm_trainer=$(func_parser_value "${lines[11]}")
|
||||
pact_trainer=$(func_parser_value "${lines[12]}")
|
||||
fpgm_trainer=$(func_parser_value "${lines[13]}")
|
||||
distill_trainer=$(func_parser_value "${lines[14]}")
|
||||
|
||||
eval_py=$(func_parser_value "${lines[15]}")
|
||||
norm_export=$(func_parser_value "${lines[16]}")
|
||||
pact_export=$(func_parser_value "${lines[17]}")
|
||||
fpgm_export=$(func_parser_value "${lines[18]}")
|
||||
distill_export=$(func_parser_value "${lines[19]}")
|
||||
|
||||
inference_py=$(func_parser_value "${lines[20]}")
|
||||
use_gpu_key=$(func_parser_key "${lines[21]}")
|
||||
use_gpu_list=$(func_parser_value "${lines[21]}")
|
||||
use_mkldnn_key=$(func_parser_key "${lines[22]}")
|
||||
use_mkldnn_list=$(func_parser_value "${lines[22]}")
|
||||
cpu_threads_key=$(func_parser_key "${lines[23]}")
|
||||
cpu_threads_list=$(func_parser_value "${lines[23]}")
|
||||
batch_size_key=$(func_parser_key "${lines[24]}")
|
||||
batch_size_list=$(func_parser_value "${lines[24]}")
|
||||
use_trt_key=$(func_parser_key "${lines[25]}")
|
||||
use_trt_list=$(func_parser_value "${lines[25]}")
|
||||
precision_key=$(func_parser_key "${lines[26]}")
|
||||
precision_list=$(func_parser_value "${lines[26]}")
|
||||
model_dir_key=$(func_parser_key "${lines[27]}")
|
||||
image_dir_key=$(func_parser_key "${lines[28]}")
|
||||
save_log_key=$(func_parser_key "${lines[29]}")
|
||||
|
||||
LOG_PATH="./test/output"
|
||||
mkdir -p ${LOG_PATH}
|
||||
status_log="${LOG_PATH}/results.log"
|
||||
|
||||
if [ ${MODE} = "lite_train_infer" ]; then
|
||||
export infer_img_dir="./train_data/icdar2015/text_localization/ch4_test_images/"
|
||||
export epoch_num=10
|
||||
elif [ ${MODE} = "whole_infer" ]; then
|
||||
export infer_img_dir="./train_data/icdar2015/text_localization/ch4_test_images/"
|
||||
export epoch_num=10
|
||||
elif [ ${MODE} = "whole_train_infer" ]; then
|
||||
export infer_img_dir="./train_data/icdar2015/text_localization/ch4_test_images/"
|
||||
export epoch_num=300
|
||||
else
|
||||
export infer_img_dir="./inference/ch_det_data_50/all-sum-510"
|
||||
export infer_model_dir="./inference/ch_ppocr_mobile_v2.0_det_train/best_accuracy"
|
||||
fi
|
||||
|
||||
|
||||
function func_inference(){
|
||||
IFS='|'
|
||||
_python=$1
|
||||
_script=$2
|
||||
_model_dir=$3
|
||||
_log_path=$4
|
||||
_img_dir=$5
|
||||
|
||||
# inference
|
||||
for use_gpu in ${use_gpu_list[*]}; do
|
||||
if [ ${use_gpu} = "False" ]; then
|
||||
for use_mkldnn in ${use_mkldnn_list[*]}; do
|
||||
for threads in ${cpu_threads_list[*]}; do
|
||||
for batch_size in ${batch_size_list[*]}; do
|
||||
_save_log_path="${_log_path}/infer_cpu_usemkldnn_${use_mkldnn}_threads_${threads}_batchsize_${batch_size}"
|
||||
command="${_python} ${_script} ${use_gpu_key}=${use_gpu} ${use_mkldnn_key}=${use_mkldnn} ${cpu_threads_key}=${threads} ${model_dir_key}=${_model_dir} ${batch_size_key}=${batch_size} ${image_dir_key}=${_img_dir} ${save_log_key}=${_save_log_path}"
|
||||
eval $command
|
||||
status_check $? "${command}" "${status_log}"
|
||||
done
|
||||
done
|
||||
done
|
||||
else
|
||||
for use_trt in ${use_trt_list[*]}; do
|
||||
for precision in ${precision_list[*]}; do
|
||||
if [ ${use_trt} = "False" ] && [ ${precision} != "fp32" ]; then
|
||||
continue
|
||||
fi
|
||||
for batch_size in ${batch_size_list[*]}; do
|
||||
_save_log_path="${_log_path}/infer_gpu_usetrt_${use_trt}_precision_${precision}_batchsize_${batch_size}"
|
||||
command="${_python} ${_script} ${use_gpu_key}=${use_gpu} ${use_trt_key}=${use_trt} ${precision_key}=${precision} ${model_dir_key}=${_model_dir} ${batch_size_key}=${batch_size} ${image_dir_key}=${_img_dir} ${save_log_key}=${_save_log_path}"
|
||||
eval $command
|
||||
status_check $? "${command}" "${status_log}"
|
||||
done
|
||||
done
|
||||
done
|
||||
fi
|
||||
done
|
||||
}
|
||||
|
||||
if [ ${MODE} != "infer" ]; then
|
||||
|
||||
IFS="|"
|
||||
for gpu in ${gpu_list[*]}; do
|
||||
use_gpu=True
|
||||
if [ ${gpu} = "-1" ];then
|
||||
use_gpu=False
|
||||
env=""
|
||||
elif [ ${#gpu} -le 1 ];then
|
||||
env="export CUDA_VISIBLE_DEVICES=${gpu}"
|
||||
elif [ ${#gpu} -le 15 ];then
|
||||
IFS=","
|
||||
array=(${gpu})
|
||||
env="export CUDA_VISIBLE_DEVICES=${array[0]}"
|
||||
IFS="|"
|
||||
else
|
||||
IFS=";"
|
||||
array=(${gpu})
|
||||
ips=${array[0]}
|
||||
gpu=${array[1]}
|
||||
IFS="|"
|
||||
fi
|
||||
for autocast in ${autocast_list[*]}; do
|
||||
for trainer in ${trainer_list[*]}; do
|
||||
if [ ${trainer} = "pact" ]; then
|
||||
run_train=${pact_trainer}
|
||||
run_export=${pact_export}
|
||||
elif [ ${trainer} = "fpgm" ]; then
|
||||
run_train=${fpgm_trainer}
|
||||
run_export=${fpgm_export}
|
||||
elif [ ${trainer} = "distill" ]; then
|
||||
run_train=${distill_trainer}
|
||||
run_export=${distill_export}
|
||||
else
|
||||
run_train=${norm_trainer}
|
||||
run_export=${norm_export}
|
||||
fi
|
||||
|
||||
if [ ${run_train} = "null" ]; then
|
||||
continue
|
||||
fi
|
||||
if [ ${run_export} = "null" ]; then
|
||||
continue
|
||||
fi
|
||||
|
||||
save_log="${LOG_PATH}/${trainer}_gpus_${gpu}_autocast_${autocast}"
|
||||
if [ ${#gpu} -le 2 ];then # epoch_num #TODO
|
||||
cmd="${python} ${run_train} ${train_use_gpu_key}=${use_gpu} ${autocast_key}=${autocast} ${epoch_key}=${epoch_num} ${save_model_key}=${save_log} "
|
||||
elif [ ${#gpu} -le 15 ];then
|
||||
cmd="${python} -m paddle.distributed.launch --gpus=${gpu} ${run_train} ${autocast_key}=${autocast} ${epoch_key}=${epoch_num} ${save_model_key}=${save_log}"
|
||||
else
|
||||
cmd="${python} -m paddle.distributed.launch --ips=${ips} --gpus=${gpu} ${run_train} ${autocast_key}=${autocast} ${epoch_key}=${epoch_num} ${save_model_key}=${save_log}"
|
||||
fi
|
||||
# run train
|
||||
eval $cmd
|
||||
status_check $? "${cmd}" "${status_log}"
|
||||
|
||||
# run eval
|
||||
eval_cmd="${python} ${eval_py} ${save_model_key}=${save_log} ${pretrain_model_key}=${save_log}/latest"
|
||||
eval $eval_cmd
|
||||
status_check $? "${eval_cmd}" "${status_log}"
|
||||
|
||||
# run export model
|
||||
save_infer_path="${save_log}"
|
||||
export_cmd="${python} ${run_export} ${save_model_key}=${save_log} ${pretrain_model_key}=${save_log}/latest ${save_infer_key}=${save_infer_path}"
|
||||
eval $export_cmd
|
||||
status_check $? "${export_cmd}" "${status_log}"
|
||||
|
||||
#run inference
|
||||
save_infer_path="${save_log}"
|
||||
func_inference "${python}" "${inference_py}" "${save_infer_path}" "${LOG_PATH}" "${infer_img_dir}"
|
||||
done
|
||||
done
|
||||
done
|
||||
|
||||
else
|
||||
save_infer_path="${LOG_PATH}/${MODE}"
|
||||
run_export=${norm_export}
|
||||
export_cmd="${python} ${run_export} ${save_model_key}=${save_infer_path} ${pretrain_model_key}=${infer_model_dir} ${save_infer_key}=${save_infer_path}"
|
||||
eval $export_cmd
|
||||
status_check $? "${export_cmd}" "${status_log}"
|
||||
|
||||
#run inference
|
||||
func_inference "${python}" "${inference_py}" "${save_infer_path}" "${LOG_PATH}" "${infer_img_dir}"
|
||||
fi
|
||||
|
|
|
|||
Loading…
Reference in New Issue