mirrors
/
PaddleClas
mirror of https://github.com/PaddlePaddle/PaddleClas.git
1
0
Fork 0
Code Issues Projects Releases Wiki Activity

support specify model_type in config 

such as norm_train and to_static_train
pull/2648/head
gaotingquan 2023-01-18 09:09:30 +00:00 committed by Tingquan Gao
parent 811b483e30
commit 2d66aeeb77
44 changed files with 237 additions and 168 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

262
test_tipc/benchmark_train.sh
View File

@ -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
# 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
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
# for log name
if [[ ${model_type} = "to_static_train" ]];then
to_static="d2sT_"
else
ln -s val_list.txt train_list.txt
to_static=""
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}/" "${FILENAME}"`
gpu_id=$(set_gpu_id $device_num)
# 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}"
# 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" $FILENAME
fi
# without profile
log_path="$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}"
# 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
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
# 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}"
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}/" "${FILENAME}"`
cmd="${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
# 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}"
# 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
done
done
done
done

1
test_tipc/configs/CSWinTransformer/CSWinTransformer_tiny_224_train_infer_python.txt
View File

@ -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==========================

7
test_tipc/configs/DenseNet/DenseNet121_train_infer_python.txt
View File

@ -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]}]

3
test_tipc/configs/GeneralRecognition/GeneralRecognition_PPLCNet_x2_5_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/GeneralRecognitionV2/GeneralRecognitionV2_PPLCNetV2_base_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/HRNet/HRNet_W18_C_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/HRNet/HRNet_W48_C_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/MobileNetV1/MobileNetV1_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/MobileNetV2/MobileNetV2_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_pact_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/MobileNetV3/MobileNetV3_large_x1_0_train_ptq_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

1
test_tipc/configs/MobileViT/MobileViT_S_train_infer_python.txt
View File

@ -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==========================

3
test_tipc/configs/PPHGNet/PPHGNet_small_train_infer_python.txt
View File

@ -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==========================

3
test_tipc/configs/PPHGNet/PPHGNet_tiny_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x0_25_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x0_35_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x0_5_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x0_75_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x1_0_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x1_5_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x2_0_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNet/PPLCNet_x2_5_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PPLCNetV2/PPLCNetV2_base_train_infer_python.txt
View File

@ -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==========================

1
test_tipc/configs/PVTV2/PVT_V2_B2_Linear_train_infer_python.txt
View File

@ -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==========================

7
test_tipc/configs/ResNeXt/ResNeXt50_32x4d_train_infer_python.txt
View File

@ -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]}]

5
test_tipc/configs/ResNet/ResNet152_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/ResNet/ResNet50_train_ampfp16_infer_python.txt
View File

@ -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

5
test_tipc/configs/ResNet/ResNet50_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/ResNet/ResNet50_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

1
test_tipc/configs/ResNet/ResNet50_train_linux_gpu_normal_amp_infer_python_linux_gpu_cpu.txt
View File

@ -54,3 +54,4 @@ null:null
batch_size:128|256
fp_items:purefp16
epoch:1
model_type:norm_train

3
test_tipc/configs/ResNet/ResNet50_train_pact_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/ResNet/ResNet50_train_ptq_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/ResNet/ResNet50_vd_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/ResNet/ResNet50_vd_train_pact_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/ResNet/ResNet50_vd_train_ptq_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/ShuffleNet/ShuffleNetV2_x1_0_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_linux_gpu_fleet_normal_infer_python_linux_gpu_cpu.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_pact_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

5
test_tipc/configs/SwinTransformer/SwinTransformer_tiny_patch4_window7_224_train_ptq_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]

3
test_tipc/configs/Twins/alt_gvt_base_train_infer_python.txt
View File

@ -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==========================

5
test_tipc/configs/Twins/alt_gvt_small_train_infer_python.txt
View File

@ -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]}]
random_infer_input:[{float32,[3,224,224]}]