提交 d666de85 编写于 作者: L LDOUBLEV

delete debug

上级 28996943
......@@ -117,7 +117,7 @@ if [ ${precision} = "null" ];then
fi
# set env
python=python3.7
python=python
export model_branch=`git symbolic-ref HEAD 2>/dev/null | cut -d"/" -f 3`
export model_commit=$(git log|head -n1|awk '{print $2}')
export str_tmp=$(echo `pip list|grep paddlepaddle-gpu|awk -F ' ' '{print $2}'`)
......@@ -129,6 +129,7 @@ export frame_commit=$(echo `${python} -c "import paddle;print(paddle.version.com
# line export_py: 30
func_sed_params "$FILENAME" "24" "null"
func_sed_params "$FILENAME" "30" "null"
func_sed_params "$FILENAME" "3" "python"
if [ ${#gpu_id} -le 1 ];then
......@@ -139,7 +140,7 @@ if [ ${#gpu_id} -le 1 ];then
# set profile_option params
IFS=";"
cmd="sed -i '13s/.*/${profile_option}/' '${FILENAME}'"
eval $cmd
eval $cmd
# run test_train_inference_python.sh
cmd="bash test_tipc/test_train_inference_python.sh ${FILENAME} benchmark_train > ${log_path}/${log_name} 2>&1 "
......@@ -165,15 +166,14 @@ if [ ${#gpu_id} -le 1 ];then
# parser log
_model_name="${model_name}_bs${batch_size}_${precision}_${run_process_type}_${run_mode}"
# cmd="${python} ${BENCHMARK_ROOT}/scripts/analysis.py --filename ${log_path}/${log_name} \
cmd="${python} analysis.py --filename ${log_path}/${log_name} \
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} \
--run_process_type ${run_process_type} \
--fp_item ${precision} \
--keyword ips: \
--keyword samples/s: \
--skip_steps 2 \
--device_num ${device_num} \
--speed_unit images/s \
......@@ -203,15 +203,14 @@ else
# parser log
_model_name="${model_name}_bs${batch_size}_${precision}_${run_process_type}_${run_mode}"
# cmd="python3.7 ${BENCHMARK_ROOT}/scripts/analysis.py --filename ${log_path}/${log_name} \
cmd="${python} analysis.py --filename ${log_path}/${log_name} \
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} \
--run_process_type ${run_process_type} \
--fp_item ${precision} \
--keyword ips: \
--keyword samples/s: \
--skip_steps 2 \
--device_num ${device_num} \
--speed_unit images/s \
......
......@@ -273,7 +273,7 @@ def train(config,
(global_step > 0 and global_step % print_batch_step == 0) or
(idx >= len(train_dataloader) - 1)):
logs = train_stats.log()
strs = 'epoch: [{}/{}], global_step: {}, {}, avg_reader_cost: {:.5f} s, avg_batch_cost: {:.5f} s, avg_samples: {}, ips: {:.5f}'.format(
strs = 'epoch: [{}/{}], global_step: {}, {}, avg_reader_cost: {:.5f} s, avg_batch_cost: {:.5f} s, avg_samples: {}, samples/s: {:.5f}'.format(
epoch, epoch_num, global_step, logs, train_reader_cost /
print_batch_step, (train_reader_cost + train_run_cost) /
print_batch_step, total_samples / print_batch_step,
......
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册