From e70d44aeddb580d21a687078cb05b4217c4ee195 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E8=8B=8F=E7=88=B1=E4=B8=9C?= <384169931@qq.com>
Date: Mon, 15 May 2023 03:48:05 +0000
Subject: [PATCH 1/2] =?UTF-8?q?=E5=88=A0=E9=99=A4=E6=96=87=E4=BB=B6=20Tens?=
 =?UTF-8?q?orFlow/built-in/recommendation/DIN=5FID0190=5Ffor=5FTensorFlow/?=
 =?UTF-8?q?test/train=5Ffull=5F8p.sh?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .../test/train_full_8p.sh                     | 160 ------------------
 1 file changed, 160 deletions(-)
 delete mode 100644 TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh

diff --git a/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh b/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh
deleted file mode 100644
index 19a169151..000000000
--- a/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh
+++ /dev/null
@@ -1,160 +0,0 @@
-#!/bin/bash
-cur_path=`pwd`
-
-#集合通信参数,不需要修改
-export RANK_SIZE=8
-export RANK_TABLE_FILE=$cur_path/rank_table_8p.json
-export JOB_ID=10087
-export OP_NO_REUSE_MEM=StridedSliceD
-
-#export ASCEND_DEVICE_ID=
-
-# 数据集路径,保持为空,不需要修改
-data_path=""
-
-#精度参数
-#precision_mode="must_keep_origin_dtype"
-
-#网络名称,同目录名称,需要模型审视修改
-Network="DIN_ID0190_for_TensorFlow"
-
-#训练batch_size,,需要模型审视修改
-batch_size=1024
-
-# 指定训练所使用的npu device卡id
-device_id=0
-# 参数校验，data_path为必传参数，其他参数的增删由模型自身决定；此处新增参数需在上面有定义并赋值
-for para in $*
-do
-    if [[ $para == --data_path* ]];then
-        data_path=`echo ${para#*=}`
-    elif [[ $para == --batch_size* ]];then
-        batch_size=`echo ${para#*=}`
-    elif [[ $para == --precision_mode* ]];then
-        precision_mode=`echo ${para#*=}`
-    elif [[ $para == --device_id* ]];then
-        device_id=`echo ${para#*=}`
-    fi
-done
-
-#校验是否传入data_path,不需要修改
-#if [[ $data_path == "" ]];then
-#    echo "[Error] para \"data_path\" must be confing"
-#    exit 1
-#fi
-# 校验是否指定了device_id,分动态分配device_id与手动指定device_id,此处不需要修改
-if [ $ASCEND_DEVICE_ID ];then
-    device_id=${ASCEND_DEVICE_ID}
-    echo "device id is ${ASCEND_DEVICE_ID}"
-elif [ ${device_id} ];then
-    export ASCEND_DEVICE_ID=${device_id}
-    echo "device id is ${ASCEND_DEVICE_ID}"
-else
-    "[Error] device id must be config"
-    exit 1
-fi
-###############指定训练脚本执行路径###############
-# cd到与test文件夹同层级目录下执行脚本，提高兼容性；test_path_dir为包含test文件夹的路径
-cur_path=`pwd`
-cur_path_last_dirname=${cur_path##*/}
-if [ x"${cur_path_last_dirname}" == x"test" ];then
-    test_path_dir=${cur_path}
-    cd ..
-    cur_path=`pwd`
-else
-    test_path_dir=${cur_path}/test
-fi
-
-#创建DeviceID输出目录，不需要修改
-if [ -d ${test_path_dir}/output/${ASCEND_DEVICE_ID} ];then
-    rm -rf ${test_path_dir}/output/$ASCEND_DEVICE_ID
-    mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
-else
-    mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
-fi
-
-#################启动训练脚本#################
-#训练开始时间，不需要修改
-start_time=$(date +%s)
-# 非平台场景时source 环境变量
-check_etp_flag=`env | grep etp_running_flag`
-etp_flag=`echo ${check_etp_flag#*=}`
-if [ x"${etp_flag}" != x"true" ];then
-    source ${test_path_dir}/env_npu.sh
-fi
-#数据集处理
-
-#执行训练脚本，以下传参不需要修改，其他需要模型审视修改
-
-cd $cur_path/examples
-sed -i  "s|./data|$data_path|g" din_demo.py
-RANK_ID_START=0
-RANK_SIZE=8
-
-for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
-do
-   #echo "Device ID: $RANK_ID"
-   export RANK_ID=$RANK_ID
-   export ASCEND_DEVICE_ID=$RANK_ID
-   ASCEND_DEVICE_ID=$RANK_ID
-  if [   -d $cur_path/test/output/${ASCEND_DEVICE_ID} ];then
-     rm -rf $cur_path/test/output/${ASCEND_DEVICE_ID}
-     mkdir -p $cur_path/test/output/${ASCEND_DEVICE_ID}
-  else
-     mkdir -p $cur_path/test/output/${ASCEND_DEVICE_ID}
-  fi
-
-    corenum=`cat /proc/cpuinfo |grep 'processor' |wc -l`
-    let a=RANK_ID*${corenum}/8
-    let b=RANK_ID+1
-    let c=b*${corenum}/8-1
-    #echo $"Device ID: $RANK_ID: "$a "-" $c
-    if [ "x${bind_core}" != x ];then
-        bind_core="taskset -c $a-$c"
-    fi
-
-  nohup  taskset -c $a-$c python3.7  din_demo.py > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
-done
-
-wait
-sed -i 's///g' ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log
-sed -i 's/
/\n/g' ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log
-##################获取训练数据################
-#训练结束时间，不需要修改
-end_time=$(date +%s)
-e2e_time=$(( $end_time - $start_time ))
-
-#结果打印，不需要修改
-echo "------------------ Final result ------------------"
-train_time=`grep -rn "val_binary_crossentropy" ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log |awk -F ' ' '{print $5}'|awk -F 'ms' '{print $1}'|awk '{sum+=$1} END {print"",sum/NR}' |awk '$1=$1'`
-FPS=`awk 'BEGIN{printf "%.2f\n", '${batch_size}'*1000/'${train_time}'}'`
-#打印，不需要修改
-echo "Final Performance images/sec : $FPS"
-#吞吐量
-
-ActualFPS=${FPS}
-BatchSize=${batch_size}
-DeviceType=`uname -m`
-CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'acc'
-TrainingTime=`awk 'BEGIN{printf "%.2f\n", '${batch_size}'*1000/'${FPS}'}'`
-
-#从train_$ASCEND_DEVICE_ID.log提取Loss到train_${CaseName}_loss.txt中，需要模型审视修改
-grep "loss" ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F 'loss: ' '{print $2}'|awk -F ' ' '{print $1}' > ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
-#最后一个迭代loss值，不需要修改
-ActualLoss=`awk 'END {print}' ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
-# 输出训练精度,需要模型审视修改
-train_accuracy=`grep -rn 'val_auc'  ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F 'val_auc: ' '{print $2}'|awk 'END {print}'`
-# 打印，不需要修改
-echo "Final Train Accuracy : ${train_accuracy}"
-
-#关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${batch_size}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainAccuracy = ${train_accuracy}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-- 
Gitee


From 7b714bbfbaedf0a4d3e01f1745e0a3e2405e31e4 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E8=8B=8F=E7=88=B1=E4=B8=9C?= <384169931@qq.com>
Date: Mon, 15 May 2023 03:48:24 +0000
Subject: [PATCH 2/2] ADD
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: 苏爱东 <384169931@qq.com>
---
 .../test/train_full_8p.sh                     | 160 ++++++++++++++++++
 .../test/train_performance_8p.sh              | 160 ++++++++++++++++++
 2 files changed, 320 insertions(+)
 create mode 100644 TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh
 create mode 100644 TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_performance_8p.sh

diff --git a/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh b/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh
new file mode 100644
index 000000000..5c63b6d7c
--- /dev/null
+++ b/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_full_8p.sh
@@ -0,0 +1,160 @@
+#!/bin/bash
+cur_path=`pwd`
+
+#集合通信参数,不需要修改
+export RANK_SIZE=8
+export RANK_TABLE_FILE=$cur_path/rank_table_8p.json
+export JOB_ID=10087
+export OP_NO_REUSE_MEM=StridedSliceD
+
+#export ASCEND_DEVICE_ID=
+
+# 数据集路径,保持为空,不需要修改
+data_path=""
+
+#精度参数
+#precision_mode="must_keep_origin_dtype"
+
+#网络名称,同目录名称,需要模型审视修改
+Network="DIN_ID0190_for_TensorFlow"
+
+#训练batch_size,,需要模型审视修改
+batch_size=1024
+
+# 指定训练所使用的npu device卡id
+device_id=0
+# 参数校验，data_path为必传参数，其他参数的增删由模型自身决定；此处新增参数需在上面有定义并赋值
+for para in $*
+do
+    if [[ $para == --data_path* ]];then
+        data_path=`echo ${para#*=}`
+    elif [[ $para == --batch_size* ]];then
+        batch_size=`echo ${para#*=}`
+    elif [[ $para == --precision_mode* ]];then
+        precision_mode=`echo ${para#*=}`
+    elif [[ $para == --device_id* ]];then
+        device_id=`echo ${para#*=}`
+    fi
+done
+
+#校验是否传入data_path,不需要修改
+#if [[ $data_path == "" ]];then
+#    echo "[Error] para \"data_path\" must be confing"
+#    exit 1
+#fi
+# 校验是否指定了device_id,分动态分配device_id与手动指定device_id,此处不需要修改
+if [ $ASCEND_DEVICE_ID ];then
+    device_id=${ASCEND_DEVICE_ID}
+    echo "device id is ${ASCEND_DEVICE_ID}"
+elif [ ${device_id} ];then
+    export ASCEND_DEVICE_ID=${device_id}
+    echo "device id is ${ASCEND_DEVICE_ID}"
+else
+    "[Error] device id must be config"
+    exit 1
+fi
+###############指定训练脚本执行路径###############
+# cd到与test文件夹同层级目录下执行脚本，提高兼容性；test_path_dir为包含test文件夹的路径
+cur_path=`pwd`
+cur_path_last_dirname=${cur_path##*/}
+if [ x"${cur_path_last_dirname}" == x"test" ];then
+    test_path_dir=${cur_path}
+    cd ..
+    cur_path=`pwd`
+else
+    test_path_dir=${cur_path}/test
+fi
+
+#创建DeviceID输出目录，不需要修改
+if [ -d ${test_path_dir}/output/${ASCEND_DEVICE_ID} ];then
+    rm -rf ${test_path_dir}/output/$ASCEND_DEVICE_ID
+    mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
+else
+    mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
+fi
+
+#################启动训练脚本#################
+#训练开始时间，不需要修改
+start_time=$(date +%s)
+# 非平台场景时source 环境变量
+check_etp_flag=`env | grep etp_running_flag`
+etp_flag=`echo ${check_etp_flag#*=}`
+if [ x"${etp_flag}" != x"true" ];then
+    source ${test_path_dir}/env_npu.sh
+fi
+#数据集处理
+
+#执行训练脚本，以下传参不需要修改，其他需要模型审视修改
+
+cd $cur_path/examples
+sed -i  "s|./data|$data_path|g" din_demo.py
+RANK_ID_START=0
+RANK_SIZE=8
+
+for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
+do
+   #echo "Device ID: $RANK_ID"
+   export RANK_ID=$RANK_ID
+   export ASCEND_DEVICE_ID=$RANK_ID
+   ASCEND_DEVICE_ID=$RANK_ID
+  if [   -d $cur_path/test/output/${ASCEND_DEVICE_ID} ];then
+     rm -rf $cur_path/test/output/${ASCEND_DEVICE_ID}
+     mkdir -p $cur_path/test/output/${ASCEND_DEVICE_ID}
+  else
+     mkdir -p $cur_path/test/output/${ASCEND_DEVICE_ID}
+  fi
+
+    corenum=`cat /proc/cpuinfo |grep 'processor' |wc -l`
+    let a=RANK_ID*${corenum}/8
+    let b=RANK_ID+1
+    let c=b*${corenum}/8-1
+    #echo $"Device ID: $RANK_ID: "$a "-" $c
+    if [ "x${bind_core}" != x ];then
+        bind_core="taskset -c $a-$c"
+    fi
+
+  nohup  taskset -c $a-$c python3.7  din_demo.py > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
+done
+
+wait
+sed -i 's///g' ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log
+sed -i 's/
/\n/g' ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log
+##################获取训练数据################
+#训练结束时间，不需要修改
+end_time=$(date +%s)
+e2e_time=$(( $end_time - $start_time ))
+
+#结果打印，不需要修改
+echo "------------------ Final result ------------------"
+train_time=`grep -rn "val_binary_crossentropy" ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log |awk -F ' ' '{print $5}'|awk -F 'ms' '{print $1}'|awk '{sum+=$1} END {print"",sum/NR}' |awk '$1=$1'`
+FPS=`awk 'BEGIN{printf "%.2f\n", '${batch_size}'*1000*'$RANK_SIZE'/'${train_time}'}'`
+#打印，不需要修改
+echo "Final Performance images/sec : $FPS"
+#吞吐量
+
+ActualFPS=${FPS}
+BatchSize=${batch_size}
+DeviceType=`uname -m`
+CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'acc'
+TrainingTime=`awk 'BEGIN{printf "%.2f\n", '${batch_size}'*1000/'${FPS}'}'`
+
+#从train_$ASCEND_DEVICE_ID.log提取Loss到train_${CaseName}_loss.txt中，需要模型审视修改
+grep "loss" ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F 'loss: ' '{print $2}'|awk -F ' ' '{print $1}' > ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
+#最后一个迭代loss值，不需要修改
+ActualLoss=`awk 'END {print}' ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
+# 输出训练精度,需要模型审视修改
+train_accuracy=`grep -rn 'val_auc'  ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F 'val_auc: ' '{print $2}'|awk 'END {print}'`
+# 打印，不需要修改
+echo "Final Train Accuracy : ${train_accuracy}"
+
+#关键信息打印到${CaseName}.log中，不需要修改
+echo "Network = ${Network}" > ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "RankSize = ${RANK_SIZE}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "BatchSize = ${batch_size}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "DeviceType = ${DeviceType}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "CaseName = ${CaseName}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "ActualFPS = ${ActualFPS}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "TrainAccuracy = ${train_accuracy}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "TrainingTime = ${TrainingTime}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "ActualLoss = ${ActualLoss}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "E2ETrainingTime = ${e2e_time}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
diff --git a/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_performance_8p.sh b/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_performance_8p.sh
new file mode 100644
index 000000000..2f8698292
--- /dev/null
+++ b/TensorFlow/built-in/recommendation/DIN_ID0190_for_TensorFlow/test/train_performance_8p.sh
@@ -0,0 +1,160 @@
+#!/bin/bash
+cur_path=`pwd`
+
+#集合通信参数,不需要修改
+export RANK_SIZE=8
+export RANK_TABLE_FILE=$cur_path/rank_table_8p.json
+export JOB_ID=10087
+export OP_NO_REUSE_MEM=StridedSliceD
+
+#export ASCEND_DEVICE_ID=
+
+# 数据集路径,保持为空,不需要修改
+data_path=""
+
+#精度参数
+#precision_mode="must_keep_origin_dtype"
+
+#网络名称,同目录名称,需要模型审视修改
+Network="DIN_ID0190_for_TensorFlow"
+
+#训练batch_size,,需要模型审视修改
+batch_size=1024
+
+# 指定训练所使用的npu device卡id
+device_id=0
+# 参数校验，data_path为必传参数，其他参数的增删由模型自身决定；此处新增参数需在上面有定义并赋值
+for para in $*
+do
+    if [[ $para == --data_path* ]];then
+        data_path=`echo ${para#*=}`
+    elif [[ $para == --batch_size* ]];then
+        batch_size=`echo ${para#*=}`
+    elif [[ $para == --precision_mode* ]];then
+        precision_mode=`echo ${para#*=}`
+    elif [[ $para == --device_id* ]];then
+        device_id=`echo ${para#*=}`
+    fi
+done
+
+#校验是否传入data_path,不需要修改
+#if [[ $data_path == "" ]];then
+#    echo "[Error] para \"data_path\" must be confing"
+#    exit 1
+#fi
+# 校验是否指定了device_id,分动态分配device_id与手动指定device_id,此处不需要修改
+if [ $ASCEND_DEVICE_ID ];then
+    device_id=${ASCEND_DEVICE_ID}
+    echo "device id is ${ASCEND_DEVICE_ID}"
+elif [ ${device_id} ];then
+    export ASCEND_DEVICE_ID=${device_id}
+    echo "device id is ${ASCEND_DEVICE_ID}"
+else
+    "[Error] device id must be config"
+    exit 1
+fi
+###############指定训练脚本执行路径###############
+# cd到与test文件夹同层级目录下执行脚本，提高兼容性；test_path_dir为包含test文件夹的路径
+cur_path=`pwd`
+cur_path_last_dirname=${cur_path##*/}
+if [ x"${cur_path_last_dirname}" == x"test" ];then
+    test_path_dir=${cur_path}
+    cd ..
+    cur_path=`pwd`
+else
+    test_path_dir=${cur_path}/test
+fi
+
+#创建DeviceID输出目录，不需要修改
+if [ -d ${test_path_dir}/output/${ASCEND_DEVICE_ID} ];then
+    rm -rf ${test_path_dir}/output/$ASCEND_DEVICE_ID
+    mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
+else
+    mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
+fi
+
+#################启动训练脚本#################
+#训练开始时间，不需要修改
+start_time=$(date +%s)
+# 非平台场景时source 环境变量
+check_etp_flag=`env | grep etp_running_flag`
+etp_flag=`echo ${check_etp_flag#*=}`
+if [ x"${etp_flag}" != x"true" ];then
+    source ${test_path_dir}/env_npu.sh
+fi
+#数据集处理
+
+#执行训练脚本，以下传参不需要修改，其他需要模型审视修改
+
+cd $cur_path/examples
+sed -i  "s|./data|$data_path|g" din_demo.py
+RANK_ID_START=0
+RANK_SIZE=8
+
+for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
+do
+   #echo "Device ID: $RANK_ID"
+   export RANK_ID=$RANK_ID
+   export ASCEND_DEVICE_ID=$RANK_ID
+   ASCEND_DEVICE_ID=$RANK_ID
+  if [   -d $cur_path/test/output/${ASCEND_DEVICE_ID} ];then
+     rm -rf $cur_path/test/output/${ASCEND_DEVICE_ID}
+     mkdir -p $cur_path/test/output/${ASCEND_DEVICE_ID}
+  else
+     mkdir -p $cur_path/test/output/${ASCEND_DEVICE_ID}
+  fi
+
+    corenum=`cat /proc/cpuinfo |grep 'processor' |wc -l`
+    let a=RANK_ID*${corenum}/8
+    let b=RANK_ID+1
+    let c=b*${corenum}/8-1
+    #echo $"Device ID: $RANK_ID: "$a "-" $c
+    if [ "x${bind_core}" != x ];then
+        bind_core="taskset -c $a-$c"
+    fi
+
+  nohup  taskset -c $a-$c python3.7  din_demo.py > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
+done
+
+wait
+sed -i 's///g' ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log
+sed -i 's/
/\n/g' ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log
+##################获取训练数据################
+#训练结束时间，不需要修改
+end_time=$(date +%s)
+e2e_time=$(( $end_time - $start_time ))
+
+#结果打印，不需要修改
+echo "------------------ Final result ------------------"
+train_time=`grep -rn "val_binary_crossentropy" ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log |awk -F ' ' '{print $5}'|awk -F 'ms' '{print $1}'|awk '{sum+=$1} END {print"",sum/NR}' |awk '$1=$1'`
+FPS=`awk 'BEGIN{printf "%.2f\n", '${batch_size}'*1000*'$RANK_SIZE'/'${train_time}'}'`
+#打印，不需要修改
+echo "Final Performance images/sec : $FPS"
+#吞吐量
+
+ActualFPS=${FPS}
+BatchSize=${batch_size}
+DeviceType=`uname -m`
+CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'perf'
+TrainingTime=`awk 'BEGIN{printf "%.2f\n", '${batch_size}'*1000/'${FPS}'}'`
+
+#从train_$ASCEND_DEVICE_ID.log提取Loss到train_${CaseName}_loss.txt中，需要模型审视修改
+grep "loss" ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F 'loss: ' '{print $2}'|awk -F ' ' '{print $1}' > ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
+#最后一个迭代loss值，不需要修改
+ActualLoss=`awk 'END {print}' ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
+# 输出训练精度,需要模型审视修改
+#train_accuracy=`grep -rn 'val_auc'  ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F 'val_auc: ' '{print $2}'|awk 'END {print}'`
+# 打印，不需要修改
+echo "Final Train Accuracy : ${train_accuracy}"
+
+#关键信息打印到${CaseName}.log中，不需要修改
+echo "Network = ${Network}" > ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "RankSize = ${RANK_SIZE}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "BatchSize = ${batch_size}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "DeviceType = ${DeviceType}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "CaseName = ${CaseName}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "ActualFPS = ${ActualFPS}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+#echo "TrainAccuracy = ${train_accuracy}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "TrainingTime = ${TrainingTime}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "ActualLoss = ${ActualLoss}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
+echo "E2ETrainingTime = ${e2e_time}" >> ${test_path_dir}/output/$ASCEND_DEVICE_ID/${CaseName}.log
-- 
Gitee