From 54a4a54cdede2f3c42b8953b6f91af61f9fb1c21 Mon Sep 17 00:00:00 2001
From: Bailang <bailang12@h-partners.com>
Date: Mon, 9 May 2022 15:06:38 +0800
Subject: [PATCH 1/5] =?UTF-8?q?[=E8=87=AA=E7=A0=94][PyTorch]EfficientNet?=
 =?UTF-8?q?=5Ffor=5FPyTorch=E6=A8=A1=E5=9E=8B=E8=AE=AD=E7=BB=83=E5=90=AF?=
 =?UTF-8?q?=E5=8A=A8=E8=84=9A=E6=9C=AC=E5=8F=AA=E7=95=99test=E4=B8=8Bshell?=
 =?UTF-8?q?=E8=84=9A=E6=9C=AC=EF=BC=8C=E4=BC=98=E5=8C=96test=E4=B8=8B?=
 =?UTF-8?q?=E8=84=9A=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Bailang <bailang12@h-partners.com>
---
 .../EfficientNet_for_PyTorch/README.md        |  22 +-
 .../EfficientNet_for_PyTorch/run_1p.sh        |  42 --
 .../EfficientNet_for_PyTorch/run_8p.sh        |  39 --
 .../{ => test}/env_npu.sh                     |   0
 .../test/train_full_8p.sh                     | 364 +++++++++---------
 .../test/train_performance_1p.sh              | 335 ++++++++--------
 .../test/train_performance_8p.sh              | 362 ++++++++---------
 7 files changed, 563 insertions(+), 601 deletions(-)
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_1p.sh
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_8p.sh
 rename PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/{ => test}/env_npu.sh (100%)

diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/README.md b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/README.md
index c96358198e..8c8dbfb5be 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/README.md
+++ b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/README.md
@@ -2,22 +2,18 @@
     
 单卡训练流程：
 
-```
-	1.安装环境
-	2.修改run_1p.sh字段"data"为当前磁盘的数据集路径
-	3.修改字段device_id（单卡训练所使用的device id），为训练配置device_id，比如device_id=0
-	4.cd到run_1p.sh文件的目录，执行bash run_1p.sh单卡脚本， 进行单卡训练
-```
+    1.安装环境  
+    2.开始训练
+        bash ./test/train_full_1p.sh  --data_path=数据集路径            # 精度训练
+        bash ./test/train_performance_1p.sh  --data_path=数据集路径    # 性能训练
 
-	
+ 
 多卡训练流程
 
-```
-	1.安装环境
-	2.修改多P脚本中字段"data"为当前磁盘的数据集路径
-	3.修改run_8p.sh字段"addr"为当前主机ip地址
-	4.cd到run_8p.sh文件的目录，执行bash run_8p.sh等多卡脚本， 进行多卡训练	
-```
+    1.安装环境
+    2.开始训练
+        bash ./test/train_full_8p.sh  --data_path=数据集路径          # 精度训练
+        bash ./test/train_performance_8p.sh  --data_path=数据集路径     # 性能训练
 
 
 
diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_1p.sh
deleted file mode 100644
index ad12e922b6..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_1p.sh
+++ /dev/null
@@ -1,42 +0,0 @@
-#!/usr/bin/env bash
-source env_npu.sh
-
-/usr/local/Ascend/driver/tools/msnpureport -d 0 -g error
-
-currentDir=$(cd "$(dirname "$0")";pwd)
-currtime=`date +%Y%m%d%H%M%S`
-train_log_dir=${currentDir}/result/training_1p_job_${currtime}
-mkdir -p ${train_log_dir}
-cd ${train_log_dir}
-echo "train log path is ${train_log_dir}"
-
-if [ $(uname -m) = "aarch64" ]
-then
-	taskset -c 0-95 python3.7 -u ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=512 \
-    --lr=0.2 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --npu=0 > ${train_log_dir}/train_1p.log 2>&1 &
-else
-    python3.7 -u ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=512 \
-    --lr=0.2 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --npu=0 > ${train_log_dir}/train_1p.log 2>&1 &
-fi
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_8p.sh
deleted file mode 100644
index 2a1b9e2c31..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/run_8p.sh
+++ /dev/null
@@ -1,39 +0,0 @@
-#!/usr/bin/env bash
-source env_npu.sh
-
-/usr/local/Ascend/driver/tools/msnpureport -d 0 -g error
-/usr/local/Ascend/driver/tools/msnpureport -d 4 -g error
-
-currentDir=$(cd "$(dirname "$0")";pwd)
-currtime=`date +%Y%m%d%H%M%S`
-train_log_dir=${currentDir}/result/training_8p_job_${currtime}
-mkdir -p ${train_log_dir}
-cd ${train_log_dir}
-echo "train log path is ${train_log_dir}"
-
-kernel_num=$(nproc)
-
-if [ ${kernel_num} -lt 95 ];then
-    cpu_number=${kernel_num}
-else
-    cpu_number=95
-fi
-
-taskset -c 0-${cpu_number} python3.7 ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=4096 \
-    --lr=1.6 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --addr=$(hostname -I |awk '{print $1}') \
-    --dist-backend=hccl \
-    --multiprocessing-distributed \
-    --world-size 1 \
-    --rank 0 \
-    --device_list '0,1,2,3,4,5,6,7' > ${train_log_dir}/train_8p.log 2>&1 &
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/env_npu.sh b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/env_npu.sh
similarity index 100%
rename from PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/env_npu.sh
rename to PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/env_npu.sh
diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_full_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_full_8p.sh
index 5a841a3b88..804ed46b6e 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_full_8p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_full_8p.sh
@@ -1,174 +1,190 @@
-#!/bin/bash
-
-#当前路径,不需要修改
-cur_path=`pwd`
-#export ASCEND_SLOG_PRINT_TO_STDOUT=1
-
-#集合通信参数,不需要修改
-export HCCL_WHITELIST_DISABLE=1
-export RANK_SIZE=8
-export JOB_ID=10087
-RANK_ID_START=0
-# source env.sh
-RANK_SIZE=8
-# 数据集路径,保持为空,不需要修改
-data_path=""
-
-#设置默认日志级别,不需要修改
-# export ASCEND_GLOBAL_LOG_LEVEL_ETP_ETP=3
-
-#基础参数，需要模型审视修改
-#网络名称，同目录名称
-Network="EfficientNet_ID0097_for_PyTorch"
-#训练epoch
-train_epochs=100
-#训练batch_size
-batch_size=4096
-#训练step
-train_steps=`expr 1281167 / ${batch_size}`
-#学习率
-learning_rate=0.045
-
-
-
-#维测参数，precision_mode需要模型审视修改
-precision_mode="allow_mix_precision"
-#维持参数，以下不需要修改
-over_dump=False
-data_dump_flag=False
-data_dump_step="10"
-profiling=False
-
-
-if [[ $1 == --help || $1 == --h ]];then
-   echo "usage:./train_performance_1p.sh --data_path=data_dir --batch_size=1024 --learning_rate=0.04"
-   exit 1
-fi
-
-for para in $*
-do
-    if [[ $para == --data_path* ]];then
-      data_path=`echo ${para#*=}`
-    elif [[ $para == --batch_size* ]];then
-      batch_size=`echo ${para#*=}`
-    elif [[ $para == --learning_rate* ]];then
-      learning_rate=`echo ${para#*=}`
-    elif [[ $para == --precision_mode* ]];then
-        precision_mode=`echo ${para#*=}`
-    fi
-done
-
-PREC=""
-if [[ $precision_mode == "amp" ]];then
-  PREC="--amp"
-fi
-
-#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
-
-cd $cur_path
-
-#add ASCEND_DEVICE_ID
-ASCEND_DEVICE_ID=0
-
-#设置环境变量，不需要修改
-echo "Device ID: $ASCEND_DEVICE_ID"
-export RANK_ID=$RANK_ID
-
-if [ -d $cur_path/output ];then
-   rm -rf $cur_path/output/*
-   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
-else
-   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
-fi
-wait
-
-#修改参数
-sed -i "s|pass|break|g" ${cur_path}/../examples/imagenet/main.py
-wait
-
-#训练开始时间，不需要修改
-start_time=$(date +%s)
-
-# 绑核，不需要的绑核的模型删除，需要模型审视修改
-corenum=`cat /proc/cpuinfo |grep "processor"|wc -l`
-let a=RANK_ID*${corenum}/${RANK_SIZE}
-let b=RANK_ID+1
-let c=b*${corenum}/${RANK_SIZE}-1
- 
-
-nohup taskset -c $a-$c python3.7 ${cur_path}/../examples/imagenet/main.py \
-    --data=$data_path \
-    --arch=efficientnet-b0 --batch-size=$batch_size \
-    --lr=1.6 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --addr=$(hostname -I |awk '{print $1}') \
-    --dist-backend=hccl \
-    --multiprocessing-distributed \
-    --world-size 1 \
-    --rank 0 > $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log 2>&1 &
-wait
-
-#训练结束时间，不需要修改
-end_time=$(date +%s)
-e2e_time=$(( $end_time - $start_time ))
-
-#参数改回
-#修改参数
-sed -i "s|break|pass|g" ${cur_path}/../examples/imagenet/main.py
-wait
-
-#结果打印，不需要修改
-echo "------------------ Final result ------------------"
-#输出性能FPS，需要模型审视修改
-FPS=`grep FPS ${cur_path}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|tail -1|awk '{print $NF}'`
-
-#打印，不需要修改
-echo "Final Performance images/sec : $FPS"
-
-#输出训练精度,需要模型审视修改
-#train_accuracy=`grep -a '* Acc@1' train_0.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
-train_accuracy=`grep -a '* Acc@1' ${cur_path}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
-
-#打印，不需要修改
-#echo "Final Train Accuracy : ${train_accuracy}"
-echo "E2E Training Duration sec : $e2e_time"
-
-#性能看护结果汇总
-#训练用例信息，不需要修改
-BatchSize=${batch_size}
-DeviceType=`uname -m`
-CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'acc'
-
-##获取性能数据，不需要修改
-#吞吐量
-ActualFPS=${FPS}
-#单迭代训练时长
-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，需要根据模型审视
-grep Epoch $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F 'Loss' '{print $2}'|awk '{print $1}' > $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
-
-#最后一个迭代loss值，不需要修改
-ActualLoss=`awk 'END {print}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
-
-#关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainAccuracy = ${train_accuracy}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
\ No newline at end of file
+#!/bin/bash
+
+#当前路径,不需要修改
+cur_path=`pwd`
+cur_path_last_dirname=${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
+
+#export ASCEND_SLOG_PRINT_TO_STDOUT=1
+
+#集合通信参数,不需要修改
+export HCCL_WHITELIST_DISABLE=1
+export RANK_SIZE=8
+export JOB_ID=10087
+RANK_ID_START=0
+# source env.sh
+RANK_SIZE=8
+# 数据集路径,保持为空,不需要修改
+data_path=""
+
+#设置默认日志级别,不需要修改
+# export ASCEND_GLOBAL_LOG_LEVEL_ETP_ETP=3
+
+#基础参数，需要模型审视修改
+#网络名称，同目录名称
+Network="EfficientNet_ID0097_for_PyTorch"
+#训练epoch
+train_epochs=100
+#训练batch_size
+batch_size=4096
+#训练step
+train_steps=`expr 1281167 / ${batch_size}`
+#学习率
+learning_rate=0.045
+
+
+
+#维测参数，precision_mode需要模型审视修改
+precision_mode="allow_mix_precision"
+#维持参数，以下不需要修改
+over_dump=False
+data_dump_flag=False
+data_dump_step="10"
+profiling=False
+
+
+if [[ $1 == --help || $1 == --h ]];then
+   echo "usage:./train_performance_1p.sh --data_path=data_dir --batch_size=1024 --learning_rate=0.04"
+   exit 1
+fi
+
+for para in $*
+do
+    if [[ $para == --data_path* ]];then
+      data_path=`echo ${para#*=}`
+    elif [[ $para == --batch_size* ]];then
+      batch_size=`echo ${para#*=}`
+    elif [[ $para == --learning_rate* ]];then
+      learning_rate=`echo ${para#*=}`
+    elif [[ $para == --precision_mode* ]];then
+        precision_mode=`echo ${para#*=}`
+    fi
+done
+
+PREC=""
+if [[ $precision_mode == "amp" ]];then
+  PREC="--amp"
+fi
+
+#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
+
+cd $cur_path
+
+#add ASCEND_DEVICE_ID
+ASCEND_DEVICE_ID=0
+
+#设置环境变量，不需要修改
+echo "Device ID: $ASCEND_DEVICE_ID"
+export RANK_ID=$RANK_ID
+
+if [ -d $cur_path/output ];then
+   rm -rf $cur_path/output/*
+   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
+else
+   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
+fi
+wait
+
+# 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
+
+#修改参数
+sed -i "s|pass|break|g" ${cur_path}/examples/imagenet/main.py
+wait
+
+#训练开始时间，不需要修改
+start_time=$(date +%s)
+
+# 绑核，不需要的绑核的模型删除，需要模型审视修改
+corenum=`cat /proc/cpuinfo |grep "processor"|wc -l`
+let a=RANK_ID*${corenum}/${RANK_SIZE}
+let b=RANK_ID+1
+let c=b*${corenum}/${RANK_SIZE}-1
+ 
+
+nohup taskset -c $a-$c python3.7 ${cur_path}/examples/imagenet/main.py \
+    --data=$data_path \
+    --arch=efficientnet-b0 --batch-size=$batch_size \
+    --lr=1.6 \
+    --momentum=0.9 \
+    --epochs=$train_epochs \
+    --autoaug \
+    --amp \
+    --pm=O1 \
+    --loss_scale=32 \
+    --val_feq=10 \
+    --addr=$(hostname -I |awk '{print $1}') \
+    --dist-backend=hccl \
+    --multiprocessing-distributed \
+    --world-size 1 \
+    --rank 0 > $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log 2>&1 &
+wait
+
+#训练结束时间，不需要修改
+end_time=$(date +%s)
+e2e_time=$(( $end_time - $start_time ))
+
+#参数改回
+#修改参数
+sed -i "s|break|pass|g" ${cur_path}/examples/imagenet/main.py
+wait
+
+#结果打印，不需要修改
+echo "------------------ Final result ------------------"
+#输出性能FPS，需要模型审视修改
+FPS=`grep FPS ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|tail -1|awk '{print $NF}'`
+
+#打印，不需要修改
+echo "Final Performance images/sec : $FPS"
+
+#输出训练精度,需要模型审视修改
+#train_accuracy=`grep -a '* Acc@1' train_0.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+train_accuracy=`grep -a '* Acc@1' ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+
+#打印，不需要修改
+#echo "Final Train Accuracy : ${train_accuracy}"
+echo "E2E Training Duration sec : $e2e_time"
+
+#性能看护结果汇总
+#训练用例信息，不需要修改
+BatchSize=${batch_size}
+DeviceType=`uname -m`
+CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'acc'
+
+##获取性能数据，不需要修改
+#吞吐量
+ActualFPS=${FPS}
+#单迭代训练时长
+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，需要根据模型审视
+grep Epoch $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F 'Loss' '{print $2}'|awk '{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_${CaseName}_loss.txt`
+
+#关键信息打印到${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 = ${BatchSize}" >> $test_path_dir/output/$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 "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
+echo "TrainAccuracy = ${train_accuracy}" >> $test_path_dir/output/$ASCEND_DEVICE_ID/${CaseName}.log
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_1p.sh
index 4a5a6d5b4c..279206e7fa 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_1p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_1p.sh
@@ -1,160 +1,175 @@
-#!/bin/bash
-
-#当前路径,不需要修改
-cur_path=`pwd`
-#export ASCEND_SLOG_PRINT_TO_STDOUT=1
-
-#集合通信参数,不需要修改
-export HCCL_WHITELIST_DISABLE=1
-export RANK_SIZE=1
-export JOB_ID=10087
-RANK_ID_START=0
-# source env.sh
-# 数据集路径,保持为空,不需要修改
-data_path=""
-
-#设置默认日志级别,不需要修改
-# export ASCEND_GLOBAL_LOG_LEVEL_ETP_ETP=3
-
-#基础参数，需要模型审视修改
-#网络名称，同目录名称
-Network="EfficientNet_ID0097_for_PyTorch"
-#训练epoch
-train_epochs=1
-#训练batch_size
-batch_size=512
-#训练step
-train_steps=`expr 1281167 / ${batch_size}`
-#学习率
-learning_rate=0.045
-
-
-
-#维测参数，precision_mode需要模型审视修改
-precision_mode="allow_mix_precision"
-#维持参数，以下不需要修改
-over_dump=False
-data_dump_flag=False
-data_dump_step="10"
-profiling=False
-
-
-if [[ $1 == --help || $1 == --h ]];then
-   echo "usage:./train_performance_1p.sh --data_path=data_dir --batch_size=1024 --learning_rate=0.04"
-   exit 1
-fi
-
-for para in $*
-do
-    if [[ $para == --data_path* ]];then
-      data_path=`echo ${para#*=}`
-    elif [[ $para == --batch_size* ]];then
-      batch_size=`echo ${para#*=}`
-    elif [[ $para == --learning_rate* ]];then
-      learning_rate=`echo ${para#*=}`
-    elif [[ $para == --precision_mode* ]];then
-        precision_mode=`echo ${para#*=}`
-    fi
-done
-
-PREC=""
-if [[ $precision_mode == "amp" ]];then
-  PREC="--amp"
-fi
-
-#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
-
-cd $cur_path
-
-#设置环境变量，不需要修改
-echo "Device ID: $ASCEND_DEVICE_ID"
-export RANK_ID=$RANK_ID
-
-if [ -d $cur_path/output ];then
-   rm -rf $cur_path/output/*
-   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
-else
-   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
-fi
-wait
-
-#修改参数
-sed -i "s|pass|break|g" ${cur_path}/../examples/imagenet/main.py
-wait
-
-#训练开始时间，不需要修改
-start_time=$(date +%s)
-
-#训练
-nohup python3.7 ${cur_path}/../examples/imagenet/main.py \
-    --data=$data_path \
-    --arch=efficientnet-b0 \
-    --batch-size=$batch_size \
-    --lr=0.2 \
-    --momentum=0.9 \
-    --epochs=$train_epochs \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --stop-step-num=1000 \
-    --npu=$ASCEND_DEVICE_ID  > $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log 2>&1 &
-wait
-
-#训练结束时间，不需要修改
-end_time=$(date +%s)
-e2e_time=$(( $end_time - $start_time ))
-
-#参数改回
-#修改参数
-sed -i "s|break|pass|g" ${cur_path}/../examples/imagenet/main.py
-wait
-
-#结果打印，不需要修改
-echo "------------------ Final result ------------------"
-#输出性能FPS，需要模型审视修改
-FPS=`grep FPS ${cur_path}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|tail -1|awk '{print $NF}'`
-
-#打印，不需要修改
-echo "Final Performance images/sec : $FPS"
-
-#输出训练精度,需要模型审视修改
-#train_accuracy=`grep -a '* Acc@1' train_0.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
-
-#打印，不需要修改
-#echo "Final Train Accuracy : ${train_accuracy}"
-echo "E2E Training Duration sec : $e2e_time"
-
-#性能看护结果汇总
-#训练用例信息，不需要修改
-BatchSize=${batch_size}
-DeviceType=`uname -m`
-CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'perf'
-
-##获取性能数据，不需要修改
-#吞吐量
-ActualFPS=${FPS}
-#单迭代训练时长
-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，需要根据模型审视
-grep Epoch $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F 'Loss' '{print $2}'|awk '{print $1}' > $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
-
-#最后一个迭代loss值，不需要修改
-ActualLoss=`awk 'END {print}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
-
-#关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
\ No newline at end of file
+#!/bin/bash
+
+#当前路径,不需要修改
+cur_path=`pwd`
+cur_path_last_dirname=${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
+#export ASCEND_SLOG_PRINT_TO_STDOUT=1
+
+#集合通信参数,不需要修改
+export HCCL_WHITELIST_DISABLE=1
+export RANK_SIZE=1
+export JOB_ID=10087
+RANK_ID_START=0
+# source env.sh
+# 数据集路径,保持为空,不需要修改
+data_path=""
+
+#设置默认日志级别,不需要修改
+# export ASCEND_GLOBAL_LOG_LEVEL_ETP_ETP=3
+
+#基础参数，需要模型审视修改
+#网络名称，同目录名称
+Network="EfficientNet_ID0097_for_PyTorch"
+#训练epoch
+train_epochs=1
+#训练batch_size
+batch_size=512
+#训练step
+train_steps=`expr 1281167 / ${batch_size}`
+#学习率
+learning_rate=0.045
+
+
+
+#维测参数，precision_mode需要模型审视修改
+precision_mode="allow_mix_precision"
+#维持参数，以下不需要修改
+over_dump=False
+data_dump_flag=False
+data_dump_step="10"
+profiling=False
+
+
+if [[ $1 == --help || $1 == --h ]];then
+   echo "usage:./train_performance_1p.sh --data_path=data_dir --batch_size=1024 --learning_rate=0.04"
+   exit 1
+fi
+
+for para in $*
+do
+    if [[ $para == --data_path* ]];then
+      data_path=`echo ${para#*=}`
+    elif [[ $para == --batch_size* ]];then
+      batch_size=`echo ${para#*=}`
+    elif [[ $para == --learning_rate* ]];then
+      learning_rate=`echo ${para#*=}`
+    elif [[ $para == --precision_mode* ]];then
+        precision_mode=`echo ${para#*=}`
+    fi
+done
+
+PREC=""
+if [[ $precision_mode == "amp" ]];then
+  PREC="--amp"
+fi
+
+#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
+
+cd $cur_path
+ASCEND_DEVICE_ID=0
+#设置环境变量，不需要修改
+echo "Device ID: $ASCEND_DEVICE_ID"
+export RANK_ID=$RANK_ID
+
+if [ -d $test_path_dir/output ];then
+   rm -rf $test_path_dir/output/*
+   mkdir -p $test_path_dir/output/$ASCEND_DEVICE_ID
+else
+   mkdir -p $test_path_dir/output/$ASCEND_DEVICE_ID
+fi
+wait
+
+# 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
+
+#修改参数
+sed -i "s|pass|break|g" ${cur_path}/examples/imagenet/main.py
+wait
+
+#训练开始时间，不需要修改
+start_time=$(date +%s)
+
+#训练
+nohup python3.7 ${cur_path}/examples/imagenet/main.py \
+    --data=$data_path \
+    --arch=efficientnet-b0 \
+    --batch-size=$batch_size \
+    --lr=0.2 \
+    --momentum=0.9 \
+    --epochs=$train_epochs \
+    --autoaug \
+    --amp \
+    --pm=O1 \
+    --loss_scale=32 \
+    --val_feq=10 \
+    --stop-step-num=1000 \
+    --npu=$ASCEND_DEVICE_ID  > $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log 2>&1 &
+wait
+
+#训练结束时间，不需要修改
+end_time=$(date +%s)
+e2e_time=$(( $end_time - $start_time ))
+
+#参数改回
+#修改参数
+sed -i "s|break|pass|g" ${cur_path}/examples/imagenet/main.py
+wait
+
+#结果打印，不需要修改
+echo "------------------ Final result ------------------"
+#输出性能FPS，需要模型审视修改
+FPS=`grep FPS ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|tail -1|awk '{print $NF}'`
+
+#打印，不需要修改
+echo "Final Performance images/sec : $FPS"
+
+#输出训练精度,需要模型审视修改
+#train_accuracy=`grep -a '* Acc@1' train_0.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+
+#打印，不需要修改
+#echo "Final Train Accuracy : ${train_accuracy}"
+echo "E2E Training Duration sec : $e2e_time"
+
+#性能看护结果汇总
+#训练用例信息，不需要修改
+BatchSize=${batch_size}
+DeviceType=`uname -m`
+CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'perf'
+
+##获取性能数据，不需要修改
+#吞吐量
+ActualFPS=${FPS}
+#单迭代训练时长
+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，需要根据模型审视
+grep Epoch $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F 'Loss' '{print $2}'|awk '{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_${CaseName}_loss.txt`
+
+#关键信息打印到${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 = ${BatchSize}" >> $test_path_dir/output/$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 "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
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_8p.sh
index 5128dfcbb2..890b892c58 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_8p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet_for_PyTorch/test/train_performance_8p.sh
@@ -1,173 +1,189 @@
-#!/bin/bash
-
-#当前路径,不需要修改
-cur_path=`pwd`
-#export ASCEND_SLOG_PRINT_TO_STDOUT=1
-export SOC_VERSION=Ascend910
-export HCCL_CONNECT_TIMEOUT=600
-
-#集合通信参数,不需要修改
-export HCCL_WHITELIST_DISABLE=1
-export RANK_SIZE=8
-export JOB_ID=10087
-RANK_ID_START=0
-# source env.sh
-RANK_SIZE=8
-# 数据集路径,保持为空,不需要修改
-data_path=""
-
-#设置默认日志级别,不需要修改
-# export ASCEND_GLOBAL_LOG_LEVEL_ETP_ETP_ETP_ETP=3
-
-#基础参数，需要模型审视修改
-#网络名称，同目录名称
-Network="EfficientNet_ID0097_for_PyTorch"
-#训练epoch
-train_epochs=1
-#训练batch_size
-batch_size=4096
-#训练step
-train_steps=`expr 1281167 / ${batch_size}`
-#学习率
-learning_rate=0.045
-
-#维测参数，precision_mode需要模型审视修改
-precision_mode="allow_mix_precision"
-#维持参数，以下不需要修改
-over_dump=False
-data_dump_flag=False
-data_dump_step="10"
-profiling=False
-
-
-if [[ $1 == --help || $1 == --h ]];then
-   echo "usage:./train_performance_1p.sh --data_path=data_dir --batch_size=1024 --learning_rate=0.04"
-   exit 1
-fi
-
-for para in $*
-do
-    if [[ $para == --data_path* ]];then
-      data_path=`echo ${para#*=}`
-    elif [[ $para == --batch_size* ]];then
-      batch_size=`echo ${para#*=}`
-    elif [[ $para == --learning_rate* ]];then
-      learning_rate=`echo ${para#*=}`
-    elif [[ $para == --precision_mode* ]];then
-        precision_mode=`echo ${para#*=}`
-    fi
-done
-
-PREC=""
-if [[ $precision_mode == "amp" ]];then
-  PREC="--amp"
-fi
-
-#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
-
-cd $cur_path
-
-#设置环境变量，不需要修改
-echo "Device ID: $ASCEND_DEVICE_ID"
-export RANK_ID=$RANK_ID
-
-if [ -d $cur_path/output ];then
-   rm -rf $cur_path/output/*
-   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
-else
-   mkdir -p $cur_path/output/$ASCEND_DEVICE_ID
-fi
-wait
-
-#修改参数
-sed -i "s|pass|break|g" ${cur_path}/../examples/imagenet/main.py
-wait
-
-#训练开始时间，不需要修改
-start_time=$(date +%s)
-
-# 绑核，不需要的绑核的模型删除，需要模型审视修改
-#corenum=`cat /proc/cpuinfo |grep "processor"|wc -l`
-#let a=RANK_ID*${corenum}/${RANK_SIZE}
-#let b=RANK_ID+1
-#let c=b*${corenum}/${RANK_SIZE}-1
-export RANK_INDEX=0
-DEVICE_INDEX=$(( ASCEND_DEVICE_ID + RANK_INDEX * 8 ))
-export DEVICE_INDEX=${DEVICE_INDEX}
-
-#nohup taskset -c $a-$c python3.7 ${cur_path}/../examples/imagenet/main.py
-nohup python3.7 ${cur_path}/../examples/imagenet/main.py \
-    --data=$data_path \
-    --arch=efficientnet-b0 \
-    --batch-size=$batch_size \
-    --lr=1.6 \
-    --momentum=0.9 \
-    --epochs=1 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --addr=$(hostname -I |awk '{print $1}') \
-    --dist-backend=hccl \
-    --multiprocessing-distributed \
-    --world-size 1 \
-    --rank 0 > $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log 2>&1 &
-wait
-
-#训练结束时间，不需要修改
-end_time=$(date +%s)
-e2e_time=$(( $end_time - $start_time ))
-
-#参数改回
-#修改参数
-sed -i "s|break|pass|g" ${cur_path}/../examples/imagenet/main.py
-wait
-
-#结果打印，不需要修改
-echo "------------------ Final result ------------------"
-#输出性能FPS，需要模型审视修改
-FPS=`grep FPS ${cur_path}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|tail -1|awk '{print $NF}'`
-
-#打印，不需要修改
-echo "Final Performance images/sec : $FPS"
-
-#输出训练精度,需要模型审视修改
-#train_accuracy=`grep -a '* Acc@1' train_0.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
-
-#打印，不需要修改
-#echo "Final Train Accuracy : ${train_accuracy}"
-echo "E2E Training Duration sec : $e2e_time"
-
-#性能看护结果汇总
-#训练用例信息，不需要修改
-BatchSize=${batch_size}
-DeviceType=`uname -m`
-CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'perf'
-
-##获取性能数据，不需要修改
-#吞吐量
-ActualFPS=${FPS}
-#单迭代训练时长
-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，需要根据模型审视
-grep Epoch $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F 'Loss' '{print $2}'|awk '{print $1}' > $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
-
-#最后一个迭代loss值，不需要修改
-ActualLoss=`awk 'END {print}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
-
-#关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
\ No newline at end of file
+#!/bin/bash
+
+#当前路径,不需要修改
+cur_path=`pwd`
+cur_path_last_dirname=${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
+
+#export ASCEND_SLOG_PRINT_TO_STDOUT=1
+export SOC_VERSION=Ascend910
+export HCCL_CONNECT_TIMEOUT=600
+
+#集合通信参数,不需要修改
+export HCCL_WHITELIST_DISABLE=1
+export RANK_SIZE=8
+export JOB_ID=10087
+RANK_ID_START=0
+# source env.sh
+RANK_SIZE=8
+# 数据集路径,保持为空,不需要修改
+data_path=""
+
+#设置默认日志级别,不需要修改
+# export ASCEND_GLOBAL_LOG_LEVEL_ETP_ETP_ETP_ETP=3
+
+#基础参数，需要模型审视修改
+#网络名称，同目录名称
+Network="EfficientNet_ID0097_for_PyTorch"
+#训练epoch
+train_epochs=1
+#训练batch_size
+batch_size=4096
+#训练step
+train_steps=`expr 1281167 / ${batch_size}`
+#学习率
+learning_rate=0.045
+
+#维测参数，precision_mode需要模型审视修改
+precision_mode="allow_mix_precision"
+#维持参数，以下不需要修改
+over_dump=False
+data_dump_flag=False
+data_dump_step="10"
+profiling=False
+
+
+if [[ $1 == --help || $1 == --h ]];then
+   echo "usage:./train_performance_1p.sh --data_path=data_dir --batch_size=1024 --learning_rate=0.04"
+   exit 1
+fi
+
+for para in $*
+do
+    if [[ $para == --data_path* ]];then
+      data_path=`echo ${para#*=}`
+    elif [[ $para == --batch_size* ]];then
+      batch_size=`echo ${para#*=}`
+    elif [[ $para == --learning_rate* ]];then
+      learning_rate=`echo ${para#*=}`
+    elif [[ $para == --precision_mode* ]];then
+        precision_mode=`echo ${para#*=}`
+    fi
+done
+
+PREC=""
+if [[ $precision_mode == "amp" ]];then
+  PREC="--amp"
+fi
+
+#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
+
+# 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
+
+cd $cur_path
+ASCEND_DEVICE_ID=0
+#设置环境变量，不需要修改
+echo "Device ID: $ASCEND_DEVICE_ID"
+export RANK_ID=$RANK_ID
+
+if [ -d $test_path_dir/output ];then
+   rm -rf $test_path_dir/output/*
+   mkdir -p $test_path_dir/output/$ASCEND_DEVICE_ID
+else
+   mkdir -p $test_path_dir/output/$ASCEND_DEVICE_ID
+fi
+wait
+
+#修改参数
+sed -i "s|pass|break|g" ${cur_path}/examples/imagenet/main.py
+wait
+
+#训练开始时间，不需要修改
+start_time=$(date +%s)
+
+# 绑核，不需要的绑核的模型删除，需要模型审视修改
+#corenum=`cat /proc/cpuinfo |grep "processor"|wc -l`
+#let a=RANK_ID*${corenum}/${RANK_SIZE}
+#let b=RANK_ID+1
+#let c=b*${corenum}/${RANK_SIZE}-1
+export RANK_INDEX=0
+DEVICE_INDEX=$(( ASCEND_DEVICE_ID + RANK_INDEX * 8 ))
+export DEVICE_INDEX=${DEVICE_INDEX}
+
+#nohup taskset -c $a-$c python3.7 ${cur_path}/../examples/imagenet/main.py
+nohup python3.7 ${cur_path}/examples/imagenet/main.py \
+    --data=$data_path \
+    --arch=efficientnet-b0 \
+    --batch-size=$batch_size \
+    --lr=1.6 \
+    --momentum=0.9 \
+    --epochs=1 \
+    --autoaug \
+    --amp \
+    --pm=O1 \
+    --loss_scale=32 \
+    --val_feq=10 \
+    --addr=$(hostname -I |awk '{print $1}') \
+    --dist-backend=hccl \
+    --multiprocessing-distributed \
+    --world-size 1 \
+    --rank 0 > $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log 2>&1 &
+wait
+
+#训练结束时间，不需要修改
+end_time=$(date +%s)
+e2e_time=$(( $end_time - $start_time ))
+
+#参数改回
+#修改参数
+sed -i "s|break|pass|g" ${cur_path}/examples/imagenet/main.py
+wait
+
+#结果打印，不需要修改
+echo "------------------ Final result ------------------"
+#输出性能FPS，需要模型审视修改
+FPS=`grep FPS ${test_path_dir}/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|tail -1|awk '{print $NF}'`
+
+#打印，不需要修改
+echo "Final Performance images/sec : $FPS"
+
+#输出训练精度,需要模型审视修改
+#train_accuracy=`grep -a '* Acc@1' train_0.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+
+#打印，不需要修改
+#echo "Final Train Accuracy : ${train_accuracy}"
+echo "E2E Training Duration sec : $e2e_time"
+
+#性能看护结果汇总
+#训练用例信息，不需要修改
+BatchSize=${batch_size}
+DeviceType=`uname -m`
+CaseName=${Network}_bs${BatchSize}_${RANK_SIZE}'p'_'perf'
+
+##获取性能数据，不需要修改
+#吞吐量
+ActualFPS=${FPS}
+#单迭代训练时长
+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，需要根据模型审视
+grep Epoch $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F 'Loss' '{print $2}'|awk '{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_${CaseName}_loss.txt`
+
+#关键信息打印到${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 = ${BatchSize}" >> $test_path_dir/output/$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 "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
\ No newline at end of file
-- 
Gitee


From 7f4d52387e63788a4236c086dc3f5c54f4b24445 Mon Sep 17 00:00:00 2001
From: Bailang <bailang12@h-partners.com>
Date: Mon, 9 May 2022 15:08:56 +0800
Subject: [PATCH 2/5] =?UTF-8?q?[=E8=87=AA=E7=A0=94][PyTorch]EfficientNet?=
 =?UTF-8?q?=5FB4=E6=A8=A1=E5=9E=8B=E8=AE=AD=E7=BB=83=E5=90=AF=E5=8A=A8?=
 =?UTF-8?q?=E8=84=9A=E6=9C=AC=E5=8F=AA=E7=95=99test=E4=B8=8Bshell=E8=84=9A?=
 =?UTF-8?q?=E6=9C=AC=EF=BC=8C=E4=BC=98=E5=8C=96test=E4=B8=8B=E8=84=9A?=
 =?UTF-8?q?=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Bailang <bailang12@h-partners.com>
---
 .../README.md                                 | 24 +++---
 .../env_npu.sh                                | 73 -------------------
 .../run_1p.sh                                 | 25 -------
 .../run_8p.sh                                 | 31 --------
 .../test/train_full_1p.sh                     | 57 +++++++++------
 .../test/train_full_8p.sh                     | 12 ++-
 .../test/train_performance_1p.sh              | 61 ++++++++++------
 .../test/train_performance_8p.sh              | 16 ++--
 8 files changed, 98 insertions(+), 201 deletions(-)
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/env_npu.sh
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_1p.sh
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_8p.sh

diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/README.md b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/README.md
index c96358198e..9325faa55c 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/README.md
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/README.md
@@ -2,26 +2,20 @@
     
 单卡训练流程：
 
-```
-	1.安装环境
-	2.修改run_1p.sh字段"data"为当前磁盘的数据集路径
-	3.修改字段device_id（单卡训练所使用的device id），为训练配置device_id，比如device_id=0
-	4.cd到run_1p.sh文件的目录，执行bash run_1p.sh单卡脚本， 进行单卡训练
-```
+    1.安装环境  
+    2.开始训练
+        bash ./test/train_full_1p.sh  --data_path=数据集路径           # 精度训练
+        bash ./test/train_performance_1p.sh  --data_path=数据集路径     # 性能训练
 
-	
+ 
 多卡训练流程
 
-```
-	1.安装环境
-	2.修改多P脚本中字段"data"为当前磁盘的数据集路径
-	3.修改run_8p.sh字段"addr"为当前主机ip地址
-	4.cd到run_8p.sh文件的目录，执行bash run_8p.sh等多卡脚本， 进行多卡训练	
-```
-
+    1.安装环境
+    2.开始训练
+        bash ./test/train_full_8p.sh  --data_path=数据集路径            # 精度训练
+        bash ./test/train_performance_8p.sh  --data_path=数据集路径      # 性能训练
 
 
-	
 二、Docker容器训练：
     
 1.导入镜像二进制包docker import ubuntuarmpytorch.tar REPOSITORY:TAG, 比如:
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/env_npu.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/env_npu.sh
deleted file mode 100644
index 1f3834e1cf..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/env_npu.sh
+++ /dev/null
@@ -1,73 +0,0 @@
-#!/bin/bash
-export install_path=/usr/local/Ascend
-
-if [ -d ${install_path}/toolkit ]; then
-    export LD_LIBRARY_PATH=/usr/include/hdf5/lib/:/usr/local/:/usr/local/lib/:/usr/lib/:${install_path}/fwkacllib/lib64/:${install_path}/driver/lib64/common/:${install_path}/driver/lib64/driver/:${install_path}/add-ons:${path_lib}:${LD_LIBRARY_PATH}
-    export PATH=${install_path}/fwkacllib/ccec_compiler/bin:${install_path}/fwkacllib/bin:$PATH
-    export PYTHONPATH=${install_path}/fwkacllib/python/site-packages:${install_path}/tfplugin/python/site-packages:${install_path}/toolkit/python/site-packages:$PYTHONPATH
-    export PYTHONPATH=/usr/local/python3.7.5/lib/python3.7/site-packages:$PYTHONPATH
-    export ASCEND_OPP_PATH=${install_path}/opp
-else
-    if [ -d ${install_path}/nnae/latest ];then
-        export LD_LIBRARY_PATH=/usr/local/:/usr/local/python3.7.5/lib/:/usr/local/openblas/lib:/usr/local/lib/:/usr/lib64/:/usr/lib/:${install_path}/nnae/latest/fwkacllib/lib64/:${install_path}/driver/lib64/common/:${install_path}/driver/lib64/driver/:${install_path}/add-ons/:/usr/lib/aarch64_64-linux-gnu:$LD_LIBRARY_PATH
-        export PATH=$PATH:${install_path}/nnae/latest/fwkacllib/ccec_compiler/bin/:${install_path}/nnae/latest/toolkit/tools/ide_daemon/bin/
-        export ASCEND_OPP_PATH=${install_path}/nnae/latest/opp/
-        export OPTION_EXEC_EXTERN_PLUGIN_PATH=${install_path}/nnae/latest/fwkacllib/lib64/plugin/opskernel/libfe.so:${install_path}/nnae/latest/fwkacllib/lib64/plugin/opskernel/libaicpu_engine.so:${install_path}/nnae/latest/fwkacllib/lib64/plugin/opskernel/libge_local_engine.so
-        export PYTHONPATH=${install_path}/nnae/latest/fwkacllib/python/site-packages/:${install_path}/nnae/latest/fwkacllib/python/site-packages/auto_tune.egg/auto_tune:${install_path}/nnae/latest/fwkacllib/python/site-packages/schedule_search.egg:$PYTHONPATH
-        export ASCEND_AICPU_PATH=${install_path}/nnae/latest
-    else
-        export LD_LIBRARY_PATH=/usr/local/:/usr/local/lib/:/usr/lib64/:/usr/lib/:/usr/local/python3.7.5/lib/:/usr/local/openblas/lib:${install_path}/ascend-toolkit/latest/fwkacllib/lib64/:${install_path}/driver/lib64/common/:${install_path}/driver/lib64/driver/:${install_path}/add-ons/:/usr/lib/aarch64-linux-gnu:$LD_LIBRARY_PATH
-        export PATH=$PATH:${install_path}/ascend-toolkit/latest/fwkacllib/ccec_compiler/bin/:${install_path}/ascend-toolkit/latest/toolkit/tools/ide_daemon/bin/
-        export ASCEND_OPP_PATH=${install_path}/ascend-toolkit/latest/opp/
-        export OPTION_EXEC_EXTERN_PLUGIN_PATH=${install_path}/ascend-toolkit/latest/fwkacllib/lib64/plugin/opskernel/libfe.so:${install_path}/ascend-toolkit/latest/fwkacllib/lib64/plugin/opskernel/libaicpu_engine.so:${install_path}/ascend-toolkit/latest/fwkacllib/lib64/plugin/opskernel/libge_local_engine.so
-        export PYTHONPATH=${install_path}/ascend-toolkit/latest/fwkacllib/python/site-packages/:${install_path}/ascend-toolkit/latest/fwkacllib/python/site-packages/auto_tune.egg/auto_tune:${install_path}/ascend-toolkit/latest/fwkacllib/python/site-packages/schedule_search.egg:$PYTHONPATH
-        export ASCEND_AICPU_PATH=${install_path}/ascend-toolkit/latest
-    fi
-fi
-
-
-#将Host日志输出到串口,0-关闭/1-开启
-export ASCEND_SLOG_PRINT_TO_STDOUT=0
-#设置默认日志级别,0-debug/1-info/2-warning/3-error
-export ASCEND_GLOBAL_LOG_LEVEL=3
-#设置Event日志开启标志,0-关闭/1-开启
-export ASCEND_GLOBAL_EVENT_ENABLE=0
-#设置是否开启taskque,0-关闭/1-开启
-export TASK_QUEUE_ENABLE=1
-#设置是否开启PTCopy,0-关闭/1-开启
-export PTCOPY_ENABLE=1
-#HCCL白名单开关,1-关闭/0-开启
-export HCCL_WHITELIST_DISABLE=1
-
-#设置device侧日志登记为error
-${install_path}/driver/tools/msnpureport -g error -d 0
-${install_path}/driver/tools/msnpureport -g error -d 1
-${install_path}/driver/tools/msnpureport -g error -d 2
-${install_path}/driver/tools/msnpureport -g error -d 3
-${install_path}/driver/tools/msnpureport -g error -d 4
-${install_path}/driver/tools/msnpureport -g error -d 5
-${install_path}/driver/tools/msnpureport -g error -d 6
-${install_path}/driver/tools/msnpureport -g error -d 7
-#关闭Device侧Event日志
-${install_path}/driver/tools/msnpureport -e disable
-
-path_lib=$(python3.7 -c """
-import sys
-import re
-result=''
-for index in range(len(sys.path)):
-    match_sit = re.search('-packages', sys.path[index])
-    if match_sit is not None:
-        match_lib = re.search('lib', sys.path[index])
-
-        if match_lib is not None:
-            end=match_lib.span()[1]
-            result += sys.path[index][0:end] + ':'
-
-        result+=sys.path[index] + '/torch/lib:'
-print(result)"""
-)
-
-echo ${path_lib}
-
-export LD_LIBRARY_PATH=/usr/local/python3.7.5/lib/:${path_lib}:$LD_LIBRARY_PATH
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_1p.sh
deleted file mode 100644
index a7bec33616..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_1p.sh
+++ /dev/null
@@ -1,25 +0,0 @@
-#!/usr/bin/env bash
-source env_npu.sh
-
-/usr/local/Ascend/driver/tools/msnpureport -d 0 -g error
-
-currentDir=$(cd "$(dirname "$0")";pwd)
-currtime=`date +%Y%m%d%H%M%S`
-train_log_dir=${currentDir}/result/training_1p_job_${currtime}
-mkdir -p ${train_log_dir}
-cd ${train_log_dir}
-echo "train log path is ${train_log_dir}"
-
-python3.7 ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=512 \
-    --lr=0.2 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --npu=0 > ${train_log_dir}/train_1p.log 2>&1 &
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_8p.sh
deleted file mode 100644
index 63e06c3410..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/run_8p.sh
+++ /dev/null
@@ -1,31 +0,0 @@
-#!/usr/bin/env bash
-source env_npu.sh
-
-/usr/local/Ascend/driver/tools/msnpureport -d 0 -g error
-/usr/local/Ascend/driver/tools/msnpureport -d 4 -g error
-
-currentDir=$(cd "$(dirname "$0")";pwd)
-currtime=`date +%Y%m%d%H%M%S`
-train_log_dir=${currentDir}/result/training_8p_job_${currtime}
-mkdir -p ${train_log_dir}
-cd ${train_log_dir}
-echo "train log path is ${train_log_dir}"
-
-python3.7 ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=4096 \
-    --lr=1.6 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --addr=$(hostname -I |awk '{print $1}') \
-    --dist-backend=hccl \
-    --multiprocessing-distributed \
-    --world-size 1 \
-    --rank 0 \
-    --device_list '0,1,2,3,4,5,6,7' > ${train_log_dir}/train_8p.log 2>&1 &
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_1p.sh
index f2f9691547..74a94c1402 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_1p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_1p.sh
@@ -2,6 +2,14 @@
 
 #当前路径,不需要修改
 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
 
 #集合通信参数,不需要修改
 
@@ -85,11 +93,18 @@ if [[ $data_path == "" ]];then
     exit 1
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-
+ASCEND_DEVICE_ID=0
 #进入训练脚本目录，需要模型审视修改
-cd $cur_path/..
+cd $cur_path
 for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
 do
     #设置环境变量，不需要修改
@@ -99,11 +114,11 @@ do
 
 
     #创建DeviceID输出目录，不需要修改
-    if [ -d ${cur_path}/output/${ASCEND_DEVICE_ID} ];then
-        rm -rf ${cur_path}/output/${ASCEND_DEVICE_ID}
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+    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/ckpt
     else
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+        mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID/ckpt
     fi
     # 绑核，不需要的绑核的模型删除，需要的模型审视修改
     #let a=RANK_ID*12
@@ -112,7 +127,7 @@ do
     
     #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
     #--data_dir, --model_dir, --precision_mode, --over_dump, --over_dump_path，--data_dump_flag，--data_dump_step，--data_dump_path，--profiling，--profiling_dump_path
-    python3 ${cur_path}/../examples/imagenet/main.py \
+    nohup python3.7 ${cur_path}/examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b4 \
         --batch-size=${batch_size} \
@@ -124,7 +139,7 @@ do
         --pm=O1 \
         --loss_scale=32 \
         --val_feq=10 \
-        --npu=$ASCEND_DEVICE_ID > ${cur_path}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
+        --npu=$ASCEND_DEVICE_ID > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
 done 
 wait
 
@@ -135,12 +150,12 @@ e2e_time=$(( $end_time - $start_time ))
 #结果打印，不需要修改
 echo "------------------ Final result ------------------"
 #输出性能FPS，需要模型审视修改
-FPS=`grep -a 'FPS'  $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
+FPS=`grep -a 'FPS'  $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
 #打印，不需要修改
 echo "Final Performance images/sec : $FPS"
 
 #输出训练精度,需要模型审视修改
-train_accuracy=`grep -a '* Acc@1' $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+train_accuracy=`grep -a '* Acc@1' $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
 #打印，不需要修改
 echo "Final Train Accuracy : ${train_accuracy}"
 echo "E2E Training Duration sec : $e2e_time"
@@ -158,18 +173,18 @@ ActualFPS=${FPS}
 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，需要根据模型审视
-grep Epoch: $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | awk -F " " '{print $1}' >> $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
+grep Epoch: $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | 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}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
+ActualLoss=`awk 'END {print}' $test_path_dir/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
 
 #关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
+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 = ${BatchSize}" >> $test_path_dir/output/$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 "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/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_8p.sh
index 6058a2132a..394afd7ec4 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_8p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_full_8p.sh
@@ -1,7 +1,4 @@
 #!/bin/bash
-#source env_npu.sh
-#当前路径,不需要修改
-cur_path=`pwd`
 
 #集合通信参数,不需要修改
 
@@ -95,12 +92,19 @@ else
     mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
 ##################启动训练脚本##################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
 # source 环境变量
 #source ${test_path_dir}/env.sh
-python3 ${test_path_dir}/../examples/imagenet/main.py \
+nohup python3.7 ${test_path_dir}/../examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b4 \
         --batch-size=${batch_size} \
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_1p.sh
index f4c1e7e92c..3e9586c3a5 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_1p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_1p.sh
@@ -1,7 +1,15 @@
 #!/bin/bash
-#source env_npu.sh
+
 #当前路径,不需要修改
 cur_path=`pwd`
+cur_path_last_dirname=${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
 
 #集合通信参数,不需要修改
 
@@ -85,25 +93,30 @@ if [[ $data_path == "" ]];then
     exit 1
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
+ASCEND_DEVICE_ID=0
 #训练开始时间，不需要修改
 start_time=$(date +%s)
 
 #进入训练脚本目录，需要模型审视修改
-cd $cur_path/..
+cd $cur_path
 for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
 do
     #设置环境变量，不需要修改
     echo "Device ID: $ASCEND_DEVICE_ID"
     export RANK_ID=$RANK_ID
-
-
-
     #创建DeviceID输出目录，不需要修改
-    if [ -d ${cur_path}/output/${ASCEND_DEVICE_ID} ];then
-        rm -rf ${cur_path}/output/${ASCEND_DEVICE_ID}
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+    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/ckpt
     else
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+        mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID/ckpt
     fi
     # 绑核，不需要的绑核的模型删除，需要的模型审视修改
     #let a=RANK_ID*12
@@ -112,7 +125,7 @@ do
     
     #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
     #--data_dir, --model_dir, --precision_mode, --over_dump, --over_dump_path，--data_dump_flag，--data_dump_step，--data_dump_path，--profiling，--profiling_dump_path
-    python3 ${cur_path}/../examples/imagenet/main.py \
+    python3 ${cur_path}/examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b4 \
         --batch-size=${batch_size} \
@@ -125,7 +138,7 @@ do
         --loss_scale=32 \
         --val_feq=10 \
 		--stop-step-num=128 \
-        --npu=$ASCEND_DEVICE_ID > ${cur_path}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
+        --npu=$ASCEND_DEVICE_ID > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
 done 
 wait
 
@@ -136,12 +149,12 @@ e2e_time=$(( $end_time - $start_time ))
 #结果打印，不需要修改
 echo "------------------ Final result ------------------"
 #输出性能FPS，需要模型审视修改
-FPS=`grep -a 'FPS'  $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
+FPS=`grep -a 'FPS'  $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
 #打印，不需要修改
 echo "Final Performance images/sec : $FPS"
 
 #输出训练精度,需要模型审视修改
-train_accuracy=`grep -a '* Acc@1' $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+train_accuracy=`grep -a '* Acc@1' $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
 #打印，不需要修改
 echo "Final Train Accuracy : ${train_accuracy}"
 echo "E2E Training Duration sec : $e2e_time"
@@ -159,18 +172,18 @@ ActualFPS=${FPS}
 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，需要根据模型审视
-grep Epoch: $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | awk -F " " '{print $1}' >> $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
+grep Epoch: $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | 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}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
+ActualLoss=`awk 'END {print}' $test_path_dir/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
 
 #关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
+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 = ${BatchSize}" >> $test_path_dir/output/$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 "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/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_8p.sh
index 23dd1394c3..6134dd7a06 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_8p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B4_ID1632_for_PyTorch/test/train_performance_8p.sh
@@ -1,10 +1,5 @@
 #!/bin/bash
-#source env_npu.sh
-#当前路径,不需要修改
-cur_path=`pwd`
-
 #集合通信参数,不需要修改
-
 export RANK_SIZE=8
 export JOB_ID=10087
 RANK_ID_START=0
@@ -95,12 +90,17 @@ else
     mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
 ##################启动训练脚本##################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-#source ${test_path_dir}/env.sh
-python3 ${test_path_dir}/../examples/imagenet/main.py \
+nohup python3.7 ${test_path_dir}/../examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b4 \
         --batch-size=${batch_size} \
-- 
Gitee


From 98a423416baf21ba038cf581b23e3511cf270272 Mon Sep 17 00:00:00 2001
From: Bailang <bailang12@h-partners.com>
Date: Mon, 9 May 2022 15:09:49 +0800
Subject: [PATCH 3/5] =?UTF-8?q?[=E8=87=AA=E7=A0=94][PyTorch]EfficientNet?=
 =?UTF-8?q?=5FB5=E6=A8=A1=E5=9E=8B=E8=AE=AD=E7=BB=83=E5=90=AF=E5=8A=A8?=
 =?UTF-8?q?=E8=84=9A=E6=9C=AC=E5=8F=AA=E7=95=99test=E4=B8=8Bshell=E8=84=9A?=
 =?UTF-8?q?=E6=9C=AC=EF=BC=8C=E4=BC=98=E5=8C=96test=E4=B8=8B=E8=84=9A?=
 =?UTF-8?q?=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Bailang <bailang12@h-partners.com>
---
 .../README.md                                 | 24 +++---
 .../env_npu.sh                                | 74 -------------------
 .../run_1p.sh                                 | 25 -------
 .../run_8p.sh                                 | 31 --------
 .../test/train_full_1p.sh                     | 59 +++++++++------
 .../test/train_full_8p.sh                     | 14 ++--
 .../test/train_performance_1p.sh              | 60 +++++++++------
 .../test/train_performance_8p.sh              | 15 ++--
 8 files changed, 99 insertions(+), 203 deletions(-)
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/env_npu.sh
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_1p.sh
 delete mode 100644 PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_8p.sh

diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/README.md b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/README.md
index c96358198e..8cfabe7a9f 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/README.md
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/README.md
@@ -1,24 +1,20 @@
 一、训练流程：
     
+
 单卡训练流程：
 
-```
-	1.安装环境
-	2.修改run_1p.sh字段"data"为当前磁盘的数据集路径
-	3.修改字段device_id（单卡训练所使用的device id），为训练配置device_id，比如device_id=0
-	4.cd到run_1p.sh文件的目录，执行bash run_1p.sh单卡脚本， 进行单卡训练
-```
+    1.安装环境  
+    2.开始训练
+        bash ./test/train_full_1p.sh  --data_path=数据集路径            # 精度训练
+        bash ./test/train_performance_1p.sh  --data_path=数据集路径    # 性能训练
 
-	
+ 
 多卡训练流程
 
-```
-	1.安装环境
-	2.修改多P脚本中字段"data"为当前磁盘的数据集路径
-	3.修改run_8p.sh字段"addr"为当前主机ip地址
-	4.cd到run_8p.sh文件的目录，执行bash run_8p.sh等多卡脚本， 进行多卡训练	
-```
-
+    1.安装环境
+    2.开始训练
+        bash ./test/train_full_8p.sh  --data_path=数据集路径          # 精度训练
+        bash ./test/train_performance_8p.sh  --data_path=数据集路径     # 性能训练
 
 
 	
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/env_npu.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/env_npu.sh
deleted file mode 100644
index b10494a71a..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/env_npu.sh
+++ /dev/null
@@ -1,74 +0,0 @@
-#!/bin/bash
-export install_path=/usr/local/Ascend
-
-if [ -d ${install_path}/toolkit ]; then
-    export LD_LIBRARY_PATH=/usr/include/hdf5/lib/:/usr/local/:/usr/local/lib/:/usr/lib/:${install_path}/fwkacllib/lib64/:${install_path}/driver/lib64/common/:${install_path}/driver/lib64/driver/:${install_path}/add-ons:${path_lib}:${LD_LIBRARY_PATH}
-    export PATH=${install_path}/fwkacllib/ccec_compiler/bin:${install_path}/fwkacllib/bin:$PATH
-    export PYTHONPATH=${install_path}/fwkacllib/python/site-packages:${install_path}/tfplugin/python/site-packages:${install_path}/toolkit/python/site-packages:$PYTHONPATH
-    export PYTHONPATH=/usr/local/python3.7.5/lib/python3.7/site-packages:$PYTHONPATH
-    export ASCEND_OPP_PATH=${install_path}/opp
-else
-    if [ -d ${install_path}/nnae/latest ];then
-        export LD_LIBRARY_PATH=/usr/local/:/usr/local/python3.7.5/lib/:/usr/local/openblas/lib:/usr/local/lib/:/usr/lib64/:/usr/lib/:${install_path}/nnae/latest/fwkacllib/lib64/:${install_path}/driver/lib64/common/:${install_path}/driver/lib64/driver/:${install_path}/add-ons/:/usr/lib/aarch64_64-linux-gnu:$LD_LIBRARY_PATH
-        export PATH=$PATH:${install_path}/nnae/latest/fwkacllib/ccec_compiler/bin/:${install_path}/nnae/latest/toolkit/tools/ide_daemon/bin/
-        export ASCEND_OPP_PATH=${install_path}/nnae/latest/opp/
-        export OPTION_EXEC_EXTERN_PLUGIN_PATH=${install_path}/nnae/latest/fwkacllib/lib64/plugin/opskernel/libfe.so:${install_path}/nnae/latest/fwkacllib/lib64/plugin/opskernel/libaicpu_engine.so:${install_path}/nnae/latest/fwkacllib/lib64/plugin/opskernel/libge_local_engine.so
-        export PYTHONPATH=${install_path}/nnae/latest/fwkacllib/python/site-packages/:${install_path}/nnae/latest/fwkacllib/python/site-packages/auto_tune.egg/auto_tune:${install_path}/nnae/latest/fwkacllib/python/site-packages/schedule_search.egg:$PYTHONPATH
-        export ASCEND_AICPU_PATH=${install_path}/nnae/latest
-    else
-        export LD_LIBRARY_PATH=/usr/local/:/usr/local/lib/:/usr/lib64/:/usr/lib/:/usr/local/python3.7.5/lib/:/usr/local/openblas/lib:${install_path}/ascend-toolkit/latest/fwkacllib/lib64/:${install_path}/driver/lib64/common/:${install_path}/driver/lib64/driver/:${install_path}/add-ons/:/usr/lib/aarch64-linux-gnu:$LD_LIBRARY_PATH
-        export PATH=$PATH:${install_path}/ascend-toolkit/latest/fwkacllib/ccec_compiler/bin/:${install_path}/ascend-toolkit/latest/toolkit/tools/ide_daemon/bin/
-        export ASCEND_OPP_PATH=${install_path}/ascend-toolkit/latest/opp/
-        export OPTION_EXEC_EXTERN_PLUGIN_PATH=${install_path}/ascend-toolkit/latest/fwkacllib/lib64/plugin/opskernel/libfe.so:${install_path}/ascend-toolkit/latest/fwkacllib/lib64/plugin/opskernel/libaicpu_engine.so:${install_path}/ascend-toolkit/latest/fwkacllib/lib64/plugin/opskernel/libge_local_engine.so
-        export PYTHONPATH=${install_path}/ascend-toolkit/latest/fwkacllib/python/site-packages/:${install_path}/ascend-toolkit/latest/fwkacllib/python/site-packages/auto_tune.egg/auto_tune:${install_path}/ascend-toolkit/latest/fwkacllib/python/site-packages/schedule_search.egg:$PYTHONPATH
-        export ASCEND_AICPU_PATH=${install_path}/ascend-toolkit/latest
-    fi
-fi
-
-
-#将Host日志输出到串口,0-关闭/1-开启
-export ASCEND_SLOG_PRINT_TO_STDOUT=0
-#设置默认日志级别,0-debug/1-info/2-warning/3-error
-export ASCEND_GLOBAL_LOG_LEVEL=3
-#设置Event日志开启标志,0-关闭/1-开启
-export ASCEND_GLOBAL_EVENT_ENABLE=0
-#设置是否开启taskque,0-关闭/1-开启
-export TASK_QUEUE_ENABLE=1
-#设置是否开启PTCopy,0-关闭/1-开启
-export PTCOPY_ENABLE=1
-#HCCL白名单开关,1-关闭/0-开启
-export HCCL_WHITELIST_DISABLE=1
-
-
-#设置device侧日志登记为error
-${install_path}/driver/tools/msnpureport -g error -d 0
-${install_path}/driver/tools/msnpureport -g error -d 1
-${install_path}/driver/tools/msnpureport -g error -d 2
-${install_path}/driver/tools/msnpureport -g error -d 3
-${install_path}/driver/tools/msnpureport -g error -d 4
-${install_path}/driver/tools/msnpureport -g error -d 5
-${install_path}/driver/tools/msnpureport -g error -d 6
-${install_path}/driver/tools/msnpureport -g error -d 7
-#关闭Device侧Event日志
-${install_path}/driver/tools/msnpureport -e disable
-
-path_lib=$(python3.7 -c """
-import sys
-import re
-result=''
-for index in range(len(sys.path)):
-    match_sit = re.search('-packages', sys.path[index])
-    if match_sit is not None:
-        match_lib = re.search('lib', sys.path[index])
-
-        if match_lib is not None:
-            end=match_lib.span()[1]
-            result += sys.path[index][0:end] + ':'
-
-        result+=sys.path[index] + '/torch/lib:'
-print(result)"""
-)
-
-echo ${path_lib}
-
-export LD_LIBRARY_PATH=/usr/local/python3.7.5/lib/:${path_lib}:$LD_LIBRARY_PATH
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_1p.sh
deleted file mode 100644
index 33a92f58cd..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_1p.sh
+++ /dev/null
@@ -1,25 +0,0 @@
-#!/usr/bin/env bash
-source npu_set_env.sh
-
-/usr/local/Ascend/driver/tools/msnpureport -d 0 -g error
-
-currentDir=$(cd "$(dirname "$0")";pwd)
-currtime=`date +%Y%m%d%H%M%S`
-train_log_dir=${currentDir}/result/training_1p_job_${currtime}
-mkdir -p ${train_log_dir}
-cd ${train_log_dir}
-echo "train log path is ${train_log_dir}"
-
-python3.7 ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=512 \
-    --lr=0.2 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --npu=0 > ${train_log_dir}/train_1p.log 2>&1 &
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_8p.sh
deleted file mode 100644
index 389062c671..0000000000
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/run_8p.sh
+++ /dev/null
@@ -1,31 +0,0 @@
-#!/usr/bin/env bash
-source npu_set_env.sh
-
-/usr/local/Ascend/driver/tools/msnpureport -d 0 -g error
-/usr/local/Ascend/driver/tools/msnpureport -d 4 -g error
-
-currentDir=$(cd "$(dirname "$0")";pwd)
-currtime=`date +%Y%m%d%H%M%S`
-train_log_dir=${currentDir}/result/training_8p_job_${currtime}
-mkdir -p ${train_log_dir}
-cd ${train_log_dir}
-echo "train log path is ${train_log_dir}"
-
-python3.7 ${currentDir}/examples/imagenet/main.py \
-    --data=/data/imagenet \
-    --arch=efficientnet-b0 \
-    --batch-size=4096 \
-    --lr=1.6 \
-    --momentum=0.9 \
-    --epochs=100 \
-    --autoaug \
-    --amp \
-    --pm=O1 \
-    --loss_scale=32 \
-    --val_feq=10 \
-    --addr=$(hostname -I |awk '{print $1}') \
-    --dist-backend=hccl \
-    --multiprocessing-distributed \
-    --world-size 1 \
-    --rank 0 \
-    --device_list '0,1,2,3,4,5,6,7' > ${train_log_dir}/train_8p.log 2>&1 &
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_1p.sh
index b6a3b682b1..c542772532 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_1p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_1p.sh
@@ -2,6 +2,14 @@
 
 #当前路径,不需要修改
 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
 
 #集合通信参数,不需要修改
 
@@ -87,25 +95,30 @@ if [[ $data_path == "" ]];then
     exit 1
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
 #训练开始时间，不需要修改
 start_time=$(date +%s)
 
 #进入训练脚本目录，需要模型审视修改
-cd $cur_path/..
+cd $cur_path
+ASCEND_DEVICE_ID=0
 for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
 do
     #设置环境变量，不需要修改
     echo "Device ID: $ASCEND_DEVICE_ID"
     export RANK_ID=$RANK_ID
-
-
-
     #创建DeviceID输出目录，不需要修改
-    if [ -d ${cur_path}/output/${ASCEND_DEVICE_ID} ];then
-        rm -rf ${cur_path}/output/${ASCEND_DEVICE_ID}
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+    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/ckpt
     else
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+        mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID/ckpt
     fi
     # 绑核，不需要的绑核的模型删除，需要的模型审视修改
     #let a=RANK_ID*12
@@ -114,7 +127,7 @@ do
     
     #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
     #--data_dir, --model_dir, --precision_mode, --over_dump, --over_dump_path，--data_dump_flag，--data_dump_step，--data_dump_path，--profiling，--profiling_dump_path
-    python3 ${cur_path}/../examples/imagenet/main.py \
+    nohup python3.7 ${cur_path}/examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b5 \
         --batch-size=${batch_size} \
@@ -126,7 +139,7 @@ do
         --pm=O1 \
         --loss_scale=32 \
         --val_feq=10 \
-        --npu=$ASCEND_DEVICE_ID > ${cur_path}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
+        --npu=$ASCEND_DEVICE_ID > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
 done 
 wait
 
@@ -137,12 +150,12 @@ e2e_time=$(( $end_time - $start_time ))
 #结果打印，不需要修改
 echo "------------------ Final result ------------------"
 #输出性能FPS，需要模型审视修改
-FPS=`grep -a 'FPS'  $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
+FPS=`grep -a 'FPS'  $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
 #打印，不需要修改
 echo "Final Performance images/sec : $FPS"
 
 #输出训练精度,需要模型审视修改
-train_accuracy=`grep -a '* Acc@1' $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+train_accuracy=`grep -a '* Acc@1' $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
 #打印，不需要修改
 echo "Final Train Accuracy : ${train_accuracy}"
 echo "E2E Training Duration sec : $e2e_time"
@@ -160,18 +173,18 @@ ActualFPS=${FPS}
 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，需要根据模型审视
-grep Epoch: $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | awk -F " " '{print $1}' >> $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
+grep Epoch: $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | 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}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
+ActualLoss=`awk 'END {print}' $test_path_dir/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
 
 #关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
+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 = ${BatchSize}" >> $test_path_dir/output/$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 "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/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_8p.sh
index 9d6ee89ba2..c6c2a13397 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_8p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_full_8p.sh
@@ -1,7 +1,4 @@
 #!/bin/bash
-#source env_npu.sh
-#当前路径,不需要修改
-cur_path=`pwd`
 
 #集合通信参数,不需要修改
 
@@ -86,6 +83,12 @@ else
     test_path_dir=${cur_path}/test
 fi
 
+# 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
 
 ##################创建日志输出目录，根据模型审视##################
 # 模型采用非循环方式启动多卡训练，创建日志输出目录如下；采用循环方式启动多卡训练的模型，在循环中创建日志输出目录，可参考CRNN模型
@@ -101,9 +104,8 @@ fi
 ##################启动训练脚本##################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-#source ${test_path_dir}/env.sh
-python3 ${test_path_dir}/../examples/imagenet/main.py \
+
+nohup python3.7 ${test_path_dir}/../examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b5 \
         --batch-size=${batch_size} \
diff --git a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_1p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_1p.sh
index ac44bd886a..6eb1638724 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_1p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_1p.sh
@@ -1,7 +1,14 @@
 #!/bin/bash
-#source env_npu.sh
 #当前路径,不需要修改
 cur_path=`pwd`
+cur_path_last_dirname=${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
 
 #集合通信参数,不需要修改
 
@@ -88,25 +95,30 @@ if [[ $data_path == "" ]];then
     exit 1
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-
+ASCEND_DEVICE_ID=0
 #进入训练脚本目录，需要模型审视修改
-cd $cur_path/..
+cd $cur_path
 for((RANK_ID=$RANK_ID_START;RANK_ID<$((RANK_SIZE+RANK_ID_START));RANK_ID++));
 do
     #设置环境变量，不需要修改
     echo "Device ID: $ASCEND_DEVICE_ID"
     export RANK_ID=$RANK_ID
 
-
-
     #创建DeviceID输出目录，不需要修改
-    if [ -d ${cur_path}/output/${ASCEND_DEVICE_ID} ];then
-        rm -rf ${cur_path}/output/${ASCEND_DEVICE_ID}
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+    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/ckpt
     else
-        mkdir -p ${cur_path}/output/$ASCEND_DEVICE_ID/ckpt
+        mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID/ckpt
     fi
     # 绑核，不需要的绑核的模型删除，需要的模型审视修改
     #let a=RANK_ID*12
@@ -115,7 +127,7 @@ do
     
     #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
     #--data_dir, --model_dir, --precision_mode, --over_dump, --over_dump_path，--data_dump_flag，--data_dump_step，--data_dump_path，--profiling，--profiling_dump_path
-    python3 ${cur_path}/../examples/imagenet/main.py \
+    nohup python3.7 ${cur_path}/examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b5 \
         --batch-size=${batch_size} \
@@ -128,7 +140,7 @@ do
         --loss_scale=32 \
         --val_feq=10 \
 		--stop-step-num=128 \
-        --npu=$ASCEND_DEVICE_ID > ${cur_path}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
+        --npu=$ASCEND_DEVICE_ID > ${test_path_dir}/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log 2>&1 &
 done 
 wait
 
@@ -139,12 +151,12 @@ e2e_time=$(( $end_time - $start_time ))
 #结果打印，不需要修改
 echo "------------------ Final result ------------------"
 #输出性能FPS，需要模型审视修改
-FPS=`grep -a 'FPS'  $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
+FPS=`grep -a 'FPS'  $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk -F " " '{print $NF}'|awk 'END {print}'`
 #打印，不需要修改
 echo "Final Performance images/sec : $FPS"
 
 #输出训练精度,需要模型审视修改
-train_accuracy=`grep -a '* Acc@1' $cur_path/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
+train_accuracy=`grep -a '* Acc@1' $test_path_dir/output/${ASCEND_DEVICE_ID}/train_${ASCEND_DEVICE_ID}.log|awk 'END {print}'|awk -F "Acc@1" '{print $NF}'|awk -F " " '{print $1}'`
 #打印，不需要修改
 echo "Final Train Accuracy : ${train_accuracy}"
 echo "E2E Training Duration sec : $e2e_time"
@@ -162,18 +174,18 @@ ActualFPS=${FPS}
 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，需要根据模型审视
-grep Epoch: $cur_path/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | awk -F " " '{print $1}' >> $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt
+grep Epoch: $test_path_dir/output/$ASCEND_DEVICE_ID/train_$ASCEND_DEVICE_ID.log|awk -F "Loss" '{print $NF}' | 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}' $cur_path/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
+ActualLoss=`awk 'END {print}' $test_path_dir/output/$ASCEND_DEVICE_ID/train_${CaseName}_loss.txt`
 
 #关键信息打印到${CaseName}.log中，不需要修改
-echo "Network = ${Network}" > $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "RankSize = ${RANK_SIZE}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "BatchSize = ${BatchSize}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "DeviceType = ${DeviceType}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "CaseName = ${CaseName}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualFPS = ${ActualFPS}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "TrainingTime = ${TrainingTime}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "ActualLoss = ${ActualLoss}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
-echo "E2ETrainingTime = ${e2e_time}" >> $cur_path/output/$ASCEND_DEVICE_ID/${CaseName}.log
+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 = ${BatchSize}" >> $test_path_dir/output/$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 "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/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_8p.sh b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_8p.sh
index fdf88cce5b..a5e411f6a4 100644
--- a/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_8p.sh
+++ b/PyTorch/built-in/cv/classification/EfficientNet-B5_ID1633_for_PyTorch/test/train_performance_8p.sh
@@ -1,7 +1,4 @@
 #!/bin/bash
-#source env_npu.sh
-#当前路径,不需要修改
-cur_path=`pwd`
 
 #集合通信参数,不需要修改
 
@@ -98,12 +95,18 @@ else
     mkdir -p ${test_path_dir}/output/$ASCEND_DEVICE_ID
 fi
 
+# 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 ${test_path_dir}/env_npu.sh
+fi
+
 ##################启动训练脚本##################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-#source ${test_path_dir}/env.sh
-python3 ${test_path_dir}/../examples/imagenet/main.py \
+
+nohup python3.7 ${test_path_dir}/../examples/imagenet/main.py \
         --data=${data_path} \
         --arch=efficientnet-b5 \
         --batch-size=${batch_size} \
-- 
Gitee


From cc35fab73932401f8c2df479fc70cea59ab12883 Mon Sep 17 00:00:00 2001
From: Bailang <bailang12@h-partners.com>
Date: Mon, 9 May 2022 15:11:00 +0800
Subject: [PATCH 4/5] =?UTF-8?q?[=E8=87=AA=E7=A0=94][PyTorch]Gluon=5FResNet?=
 =?UTF-8?q?50=5Fv1c=5Ffor=5FPyTorch=E6=A8=A1=E5=9E=8B=E8=AE=AD=E7=BB=83?=
 =?UTF-8?q?=E5=90=AF=E5=8A=A8=E8=84=9A=E6=9C=AC=E5=8F=AA=E7=95=99test?=
 =?UTF-8?q?=E4=B8=8Bshell=E8=84=9A=E6=9C=AC=EF=BC=8C=E4=BC=98=E5=8C=96test?=
 =?UTF-8?q?=E4=B8=8B=E8=84=9A=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Bailang <bailang12@h-partners.com>
---
 .../Gluon_ResNet50_v1c_for_PyTorch/README.md  | 414 ++----------------
 .../README_raw.md                             | 399 +++++++++++++++++
 .../test/train_full_1p.sh                     |  12 +-
 .../test/train_full_8p.sh                     |  12 +-
 .../test/train_performance_1p.sh              |  14 +-
 .../test/train_performance_8p.sh              |  11 +-
 6 files changed, 459 insertions(+), 403 deletions(-)
 create mode 100644 PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README_raw.md

diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README.md b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README.md
index 76261cccf2..c3df4f61b9 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README.md
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README.md
@@ -1,399 +1,37 @@
-# PyTorch Image Models
-- [Sponsors](#sponsors)
-- [What's New](#whats-new)
-- [Introduction](#introduction)
-- [Models](#models)
-- [Features](#features)
-- [Results](#results)
-- [Getting Started (Documentation)](#getting-started-documentation)
-- [Train, Validation, Inference Scripts](#train-validation-inference-scripts)
-- [Awesome PyTorch Resources](#awesome-pytorch-resources)
-- [Licenses](#licenses)
-- [Citing](#citing)
+一、训练流程：
+    
 
-## Sponsors
+单卡训练流程：
 
-A big thank you to my [GitHub Sponsors](https://github.com/sponsors/rwightman) for their support!
+    1.安装环境  
+    2.开始训练
+        bash ./test/train_full_1p.sh  --data_path=数据集路径            # 精度训练
+        bash ./test/train_performance_1p.sh  --data_path=数据集路径    # 性能训练
 
-In addition to the sponsors at the link above, I've received hardware and/or cloud resources from
-* Nvidia (https://www.nvidia.com/en-us/)
-* TFRC (https://www.tensorflow.org/tfrc)
+ 
+多卡训练流程
 
-I'm fortunate to be able to dedicate significant time and money of my own supporting this and other open source projects. However, as the projects increase in scope, outside support is needed to continue with the current trajectory of hardware, infrastructure, and electricty costs.
+    1.安装环境
+    2.开始训练
+        bash ./test/train_full_8p.sh  --data_path=数据集路径          # 精度训练
+        bash ./test/train_performance_8p.sh  --data_path=数据集路径     # 性能训练
 
-## What's New
 
-### June 20, 2021
-* Release Vision Transformer 'AugReg' weights from [How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers](https://arxiv.org/abs/2106.10270)
-  * .npz weight loading support added, can load any of the 50K+ weights from the [AugReg series](https://console.cloud.google.com/storage/browser/vit_models/augreg)
-  * See [example notebook](https://colab.research.google.com/github/google-research/vision_transformer/blob/master/vit_jax_augreg.ipynb) from official impl for navigating the augreg weights
-  * Replaced all default weights w/ best AugReg variant (if possible). All AugReg 21k classifiers work.
-    * Highlights: `vit_large_patch16_384` (87.1 top-1), `vit_large_r50_s32_384` (86.2 top-1), `vit_base_patch16_384` (86.0 top-1)
-  * `vit_deit_*` renamed to just `deit_*`
-  * Remove my old small model, replace with DeiT compatible small w/ AugReg weights
-* Add 1st training of my `gmixer_24_224` MLP /w GLU, 78.1 top-1 w/ 25M params.
-* Add weights from official ResMLP release (https://github.com/facebookresearch/deit)
-* Add `eca_nfnet_l2` weights from my 'lightweight' series. 84.7 top-1 at 384x384.
-* Add distilled BiT 50x1 student and 152x2 Teacher weights from  [Knowledge distillation: A good teacher is patient and consistent](https://arxiv.org/abs/2106.05237)
-* NFNets and ResNetV2-BiT models work w/ Pytorch XLA now
-  * weight standardization uses F.batch_norm instead of std_mean (std_mean wasn't lowered)
-  * eps values adjusted, will be slight differences but should be quite close
-* Improve test coverage and classifier interface of non-conv (vision transformer and mlp) models
-* Cleanup a few classifier / flatten details for models w/ conv classifiers or early global pool
-* Please report any regressions, this PR touched quite a few models.
+	
+二、Docker容器训练：
+    
+1.导入镜像二进制包docker import ubuntuarmpytorch.tar REPOSITORY:TAG, 比如:
 
-### June 8, 2021
-* Add first ResMLP weights, trained in PyTorch XLA on TPU-VM w/ my XLA branch. 24 block variant, 79.2 top-1.
-* Add ResNet51-Q model w/ pretrained weights at 82.36 top-1.
-  * NFNet inspired block layout with quad layer stem and no maxpool
-  * Same param count (35.7M) and throughput as ResNetRS-50 but +1.5 top-1 @ 224x224 and +2.5 top-1 at 288x288
+        docker import ubuntuarmpytorch.tar pytorch:b020
 
-### May 25, 2021
-* Add LeViT, Visformer, ConViT (PR by Aman Arora), Twins (PR by paper authors) transformer models
-* Add ResMLP and gMLP MLP vision models to the existing MLP Mixer impl
-* Fix a number of torchscript issues with various vision transformer models
-* Cleanup input_size/img_size override handling and improve testing / test coverage for all vision transformer and MLP models
-* More flexible pos embedding resize (non-square) for ViT and TnT. Thanks [Alexander Soare](https://github.com/alexander-soare)
-* Add `efficientnetv2_rw_m` model and weights (started training before official code). 84.8 top-1, 53M params.
+2.执行docker_start.sh后带三个参数：步骤1生成的REPOSITORY:TAG；数据集路径；模型执行路径；比如：
 
-### May 14, 2021
-* Add EfficientNet-V2 official model defs w/ ported weights from official [Tensorflow/Keras](https://github.com/google/automl/tree/master/efficientnetv2) impl.
-  * 1k trained variants: `tf_efficientnetv2_s/m/l`
-  * 21k trained variants: `tf_efficientnetv2_s/m/l_in21k`
-  * 21k pretrained -> 1k fine-tuned: `tf_efficientnetv2_s/m/l_in21ft1k`
-  * v2 models w/ v1 scaling: `tf_efficientnetv2_b0` through `b3`
-  * Rename my prev V2 guess `efficientnet_v2s` -> `efficientnetv2_rw_s`
-  * Some blank `efficientnetv2_*` models in-place for future native PyTorch training
+        ./docker_start.sh pytorch:b020 /train/imagenet /home/Gluon_ResNet50_v1c_for_PyTorch
 
-### May 5, 2021
-* Add MLP-Mixer models and port pretrained weights from [Google JAX impl](https://github.com/google-research/vision_transformer/tree/linen)
-* Add CaiT models and pretrained weights from [FB](https://github.com/facebookresearch/deit)
-* Add ResNet-RS models and weights from [TF](https://github.com/tensorflow/tpu/tree/master/models/official/resnet/resnet_rs). Thanks [Aman Arora](https://github.com/amaarora)
-* Add CoaT models and weights. Thanks [Mohammed Rizin](https://github.com/morizin)
-* Add new ImageNet-21k weights & finetuned weights for TResNet, MobileNet-V3, ViT models. Thanks [mrT](https://github.com/mrT23)
-* Add GhostNet models and weights. Thanks [Kai Han](https://github.com/iamhankai)
-* Update ByoaNet attention modules
-   * Improve SA module inits
-   * Hack together experimental stand-alone Swin based attn module and `swinnet`
-   * Consistent '26t' model defs for experiments.
-* Add improved Efficientnet-V2S (prelim model def) weights. 83.8 top-1.
-* WandB logging support
+3.执行步骤一训练流程（环境安装除外）
+	
+三、测试结果
+    
+训练日志路径：在训练脚本的同目录下result文件夹里，如：
 
-### April 13, 2021
-* Add Swin Transformer models and weights from https://github.com/microsoft/Swin-Transformer
-
-### April 12, 2021
-* Add ECA-NFNet-L1 (slimmed down F1 w/ SiLU, 41M params) trained with this code. 84% top-1 @ 320x320. Trained at 256x256.
-* Add EfficientNet-V2S model (unverified model definition) weights. 83.3 top-1 @ 288x288. Only trained single res 224. Working on progressive training.
-* Add ByoaNet model definition (Bring-your-own-attention) w/ SelfAttention block and corresponding SA/SA-like modules and model defs
-  * Lambda Networks - https://arxiv.org/abs/2102.08602
-  * Bottleneck Transformers - https://arxiv.org/abs/2101.11605
-  * Halo Nets - https://arxiv.org/abs/2103.12731
-* Adabelief optimizer contributed by Juntang Zhuang
-
-### April 1, 2021
-* Add snazzy `benchmark.py` script for bulk `timm` model benchmarking of train and/or inference
-* Add Pooling-based Vision Transformer (PiT) models (from https://github.com/naver-ai/pit)
-  * Merged distilled variant into main for torchscript compatibility
-  * Some `timm` cleanup/style tweaks and weights have hub download support
-* Cleanup Vision Transformer (ViT) models
-  * Merge distilled (DeiT) model into main so that torchscript can work
-  * Support updated weight init (defaults to old still) that closer matches original JAX impl (possibly better training from scratch)
-  * Separate hybrid model defs into different file and add several new model defs to fiddle with, support patch_size != 1 for hybrids
-  * Fix fine-tuning num_class changes (PiT and ViT) and pos_embed resizing (Vit) with distilled variants
-  * nn.Sequential for block stack (does not break downstream compat)
-* TnT (Transformer-in-Transformer) models contributed by author (from https://gitee.com/mindspore/mindspore/tree/master/model_zoo/research/cv/TNT)
-* Add RegNetY-160 weights from DeiT teacher model
-* Add new NFNet-L0 w/ SE attn (rename `nfnet_l0b`->`nfnet_l0`) weights 82.75 top-1 @ 288x288
-* Some fixes/improvements for TFDS dataset wrapper
-
-### March 17, 2021
-* Add new ECA-NFNet-L0 (rename `nfnet_l0c`->`eca_nfnet_l0`) weights trained by myself.
-  * 82.6 top-1 @ 288x288, 82.8 @ 320x320, trained at 224x224
-  * Uses SiLU activation, approx 2x faster than `dm_nfnet_f0` and 50% faster than `nfnet_f0s` w/ 1/3 param count
-* Integrate [Hugging Face model hub](https://huggingface.co/models) into timm create_model and default_cfg handling for pretrained weight and config sharing (more on this soon!)
-* Merge HardCoRe NAS models contributed by https://github.com/yoniaflalo
-* Merge PyTorch trained EfficientNet-EL and pruned ES/EL variants contributed by [DeGirum](https://github.com/DeGirum)
-
-
-### March 7, 2021
-* First 0.4.x PyPi release w/ NFNets (& related), ByoB (GPU-Efficient, RepVGG, etc).
-* Change feature extraction for pre-activation nets (NFNets, ResNetV2) to return features before activation.
-* Tested with PyTorch 1.8 release. Updated CI to use 1.8.
-* Benchmarked several arch on RTX 3090, Titan RTX, and V100 across 1.7.1, 1.8, NGC 20.12, and 21.02. Some interesting performance variations to take note of https://gist.github.com/rwightman/bb59f9e245162cee0e38bd66bd8cd77f
-
-### Feb 18, 2021
-* Add pretrained weights and model variants for NFNet-F* models from [DeepMind Haiku impl](https://github.com/deepmind/deepmind-research/tree/master/nfnets).
-  * Models are prefixed with `dm_`. They require SAME padding conv, skipinit enabled, and activation gains applied in act fn.
-  * These models are big, expect to run out of GPU memory. With the GELU activiation + other options, they are roughly 1/2 the inference speed of my SiLU PyTorch optimized `s` variants.
-  * Original model results are based on pre-processing that is not the same as all other models so you'll see different results in the results csv (once updated).
-  * Matching the original pre-processing as closely as possible I get these results:
-    * `dm_nfnet_f6` - 86.352
-    * `dm_nfnet_f5` - 86.100
-    * `dm_nfnet_f4` - 85.834
-    * `dm_nfnet_f3` - 85.676
-    * `dm_nfnet_f2` - 85.178
-    * `dm_nfnet_f1` - 84.696
-    * `dm_nfnet_f0` - 83.464
-
-### Feb 16, 2021
-* Add Adaptive Gradient Clipping (AGC) as per https://arxiv.org/abs/2102.06171. Integrated w/ PyTorch gradient clipping via mode arg that defaults to prev 'norm' mode. For backward arg compat, clip-grad arg must be specified to enable when using train.py.
-  * AGC w/ default clipping factor `--clip-grad .01 --clip-mode agc`
-  * PyTorch global norm of 1.0 (old behaviour, always norm), `--clip-grad 1.0`
-  * PyTorch value clipping of 10, `--clip-grad 10. --clip-mode value`
-  * AGC performance is definitely sensitive to the clipping factor. More experimentation needed to determine good values for smaller batch sizes and optimizers besides those in paper. So far I've found .001-.005 is necessary for stable RMSProp training w/ NFNet/NF-ResNet.
-
-### Feb 12, 2021
-* Update Normalization-Free nets to include new NFNet-F (https://arxiv.org/abs/2102.06171) model defs
-
-### Feb 10, 2021
-* First Normalization-Free model training experiments done,
-  * nf_resnet50 - 80.68 top-1 @ 288x288, 80.31 @ 256x256
-  * nf_regnet_b1 - 79.30 @ 288x288, 78.75 @ 256x256
-* More model archs, incl a flexible ByobNet backbone ('Bring-your-own-blocks')
-  * GPU-Efficient-Networks (https://github.com/idstcv/GPU-Efficient-Networks), impl in `byobnet.py`
-  * RepVGG (https://github.com/DingXiaoH/RepVGG), impl in `byobnet.py`
-  * classic VGG (from torchvision, impl in `vgg.py`)
-* Refinements to normalizer layer arg handling and normalizer+act layer handling in some models
-* Default AMP mode changed to native PyTorch AMP instead of APEX. Issues not being fixed with APEX. Native works with `--channels-last` and `--torchscript` model training, APEX does not.
-* Fix a few bugs introduced since last pypi release
-
-### Feb 8, 2021
-* Add several ResNet weights with ECA attention. 26t & 50t trained @ 256, test @ 320. 269d train @ 256, fine-tune @320, test @ 352.
-  * `ecaresnet26t` - 79.88 top-1 @ 320x320, 79.08 @ 256x256
-  * `ecaresnet50t` - 82.35 top-1 @ 320x320, 81.52 @ 256x256
-  * `ecaresnet269d` - 84.93 top-1 @ 352x352, 84.87 @ 320x320
-* Remove separate tiered (`t`) vs tiered_narrow (`tn`) ResNet model defs, all `tn` changed to `t` and `t` models removed (`seresnext26t_32x4d` only model w/ weights that was removed).
-* Support model default_cfgs with separate train vs test resolution `test_input_size` and remove extra `_320` suffix ResNet model defs that were just for test.
-
-### Jan 30, 2021
-* Add initial "Normalization Free" NF-RegNet-B* and NF-ResNet model definitions based on [paper](https://arxiv.org/abs/2101.08692)
-
-### Jan 25, 2021
-* Add ResNetV2 Big Transfer (BiT) models w/ ImageNet-1k and 21k weights from https://github.com/google-research/big_transfer
-* Add official R50+ViT-B/16 hybrid models + weights from https://github.com/google-research/vision_transformer
-* ImageNet-21k ViT weights are added w/ model defs and representation layer (pre logits) support
-  * NOTE: ImageNet-21k classifier heads were zero'd in original weights, they are only useful for transfer learning
-* Add model defs and weights for DeiT Vision Transformer models from https://github.com/facebookresearch/deit
-* Refactor dataset classes into ImageDataset/IterableImageDataset + dataset specific parser classes
-* Add Tensorflow-Datasets (TFDS) wrapper to allow use of TFDS image classification sets with train script
-  * Ex: `train.py /data/tfds --dataset tfds/oxford_iiit_pet --val-split test --model resnet50 -b 256 --amp --num-classes 37 --opt adamw --lr 3e-4 --weight-decay .001 --pretrained -j 2`
-* Add improved .tar dataset parser that reads images from .tar, folder of .tar files, or .tar within .tar
-  * Run validation on full ImageNet-21k directly from tar w/ BiT model: `validate.py /data/fall11_whole.tar --model resnetv2_50x1_bitm_in21k --amp`
-* Models in this update should be stable w/ possible exception of ViT/BiT, possibility of some regressions with train/val scripts and dataset handling
-
-### Jan 3, 2021
-* Add SE-ResNet-152D weights
-  * 256x256 val, 0.94 crop top-1 - 83.75
-  * 320x320 val, 1.0 crop - 84.36
-* Update [results files](results/)
-
-
-## Introduction
-
-Py**T**orch **Im**age **M**odels (`timm`) is a collection of image models, layers, utilities, optimizers, schedulers, data-loaders / augmentations, and reference training / validation scripts that aim to pull together a wide variety of SOTA models with ability to reproduce ImageNet training results.
-
-The work of many others is present here. I've tried to make sure all source material is acknowledged via links to github, arxiv papers, etc in the README, documentation, and code docstrings. Please let me know if I missed anything.
-
-## Models
-
-All model architecture families include variants with pretrained weights. There are specific model variants without any weights, it is NOT a bug. Help training new or better weights is always appreciated. Here are some example [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) to get you started.
-
-A full version of the list below with source links can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/models/).
-
-* Big Transfer ResNetV2 (BiT) - https://arxiv.org/abs/1912.11370
-* Bottleneck Transformers - https://arxiv.org/abs/2101.11605
-* CaiT (Class-Attention in Image Transformers) - https://arxiv.org/abs/2103.17239
-* CoaT (Co-Scale Conv-Attentional Image Transformers) - https://arxiv.org/abs/2104.06399
-* ConViT (Soft Convolutional Inductive Biases Vision Transformers)- https://arxiv.org/abs/2103.10697
-* CspNet (Cross-Stage Partial Networks) - https://arxiv.org/abs/1911.11929
-* DeiT (Vision Transformer) - https://arxiv.org/abs/2012.12877
-* DenseNet - https://arxiv.org/abs/1608.06993
-* DLA - https://arxiv.org/abs/1707.06484
-* DPN (Dual-Path Network) - https://arxiv.org/abs/1707.01629
-* EfficientNet (MBConvNet Family)
-    * EfficientNet NoisyStudent (B0-B7, L2) - https://arxiv.org/abs/1911.04252
-    * EfficientNet AdvProp (B0-B8) - https://arxiv.org/abs/1911.09665
-    * EfficientNet (B0-B7) - https://arxiv.org/abs/1905.11946
-    * EfficientNet-EdgeTPU (S, M, L) - https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html
-    * EfficientNet V2 - https://arxiv.org/abs/2104.00298
-    * FBNet-C - https://arxiv.org/abs/1812.03443
-    * MixNet - https://arxiv.org/abs/1907.09595
-    * MNASNet B1, A1 (Squeeze-Excite), and Small - https://arxiv.org/abs/1807.11626
-    * MobileNet-V2 - https://arxiv.org/abs/1801.04381
-    * Single-Path NAS - https://arxiv.org/abs/1904.02877
-* GhostNet - https://arxiv.org/abs/1911.11907
-* gMLP - https://arxiv.org/abs/2105.08050
-* GPU-Efficient Networks - https://arxiv.org/abs/2006.14090
-* Halo Nets - https://arxiv.org/abs/2103.12731
-* HardCoRe-NAS - https://arxiv.org/abs/2102.11646
-* HRNet - https://arxiv.org/abs/1908.07919
-* Inception-V3 - https://arxiv.org/abs/1512.00567
-* Inception-ResNet-V2 and Inception-V4 - https://arxiv.org/abs/1602.07261
-* Lambda Networks - https://arxiv.org/abs/2102.08602
-* LeViT (Vision Transformer in ConvNet's Clothing) - https://arxiv.org/abs/2104.01136
-* MLP-Mixer - https://arxiv.org/abs/2105.01601
-* MobileNet-V3 (MBConvNet w/ Efficient Head) - https://arxiv.org/abs/1905.02244
-* NASNet-A - https://arxiv.org/abs/1707.07012
-* NFNet-F - https://arxiv.org/abs/2102.06171
-* NF-RegNet / NF-ResNet - https://arxiv.org/abs/2101.08692
-* PNasNet - https://arxiv.org/abs/1712.00559
-* Pooling-based Vision Transformer (PiT) - https://arxiv.org/abs/2103.16302
-* RegNet - https://arxiv.org/abs/2003.13678
-* RepVGG - https://arxiv.org/abs/2101.03697
-* ResMLP - https://arxiv.org/abs/2105.03404
-* ResNet/ResNeXt
-    * ResNet (v1b/v1.5) - https://arxiv.org/abs/1512.03385
-    * ResNeXt - https://arxiv.org/abs/1611.05431
-    * 'Bag of Tricks' / Gluon C, D, E, S variations - https://arxiv.org/abs/1812.01187
-    * Weakly-supervised (WSL) Instagram pretrained / ImageNet tuned ResNeXt101 - https://arxiv.org/abs/1805.00932
-    * Semi-supervised (SSL) / Semi-weakly Supervised (SWSL) ResNet/ResNeXts - https://arxiv.org/abs/1905.00546
-    * ECA-Net (ECAResNet) - https://arxiv.org/abs/1910.03151v4
-    * Squeeze-and-Excitation Networks (SEResNet) - https://arxiv.org/abs/1709.01507
-    * ResNet-RS - https://arxiv.org/abs/2103.07579
-* Res2Net - https://arxiv.org/abs/1904.01169
-* ResNeSt - https://arxiv.org/abs/2004.08955
-* ReXNet - https://arxiv.org/abs/2007.00992
-* SelecSLS - https://arxiv.org/abs/1907.00837
-* Selective Kernel Networks - https://arxiv.org/abs/1903.06586
-* Swin Transformer - https://arxiv.org/abs/2103.14030
-* Transformer-iN-Transformer (TNT) - https://arxiv.org/abs/2103.00112
-* TResNet - https://arxiv.org/abs/2003.13630
-* Twins (Spatial Attention in Vision Transformers) - https://arxiv.org/pdf/2104.13840.pdf
-* Vision Transformer - https://arxiv.org/abs/2010.11929
-* VovNet V2 and V1 - https://arxiv.org/abs/1911.06667
-* Xception - https://arxiv.org/abs/1610.02357
-* Xception (Modified Aligned, Gluon) - https://arxiv.org/abs/1802.02611
-* Xception (Modified Aligned, TF) - https://arxiv.org/abs/1802.02611
-
-## Features
-
-Several (less common) features that I often utilize in my projects are included. Many of their additions are the reason why I maintain my own set of models, instead of using others' via PIP:
-
-* All models have a common default configuration interface and API for
-    * accessing/changing the classifier - `get_classifier` and `reset_classifier`
-    * doing a forward pass on just the features - `forward_features` (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
-    * these makes it easy to write consistent network wrappers that work with any of the models
-* All models support multi-scale feature map extraction (feature pyramids) via create_model (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
-    * `create_model(name, features_only=True, out_indices=..., output_stride=...)`
-    * `out_indices` creation arg specifies which feature maps to return, these indices are 0 based and generally correspond to the `C(i + 1)` feature level.
-    * `output_stride` creation arg controls output stride of the network by using dilated convolutions. Most networks are stride 32 by default. Not all networks support this.
-    * feature map channel counts, reduction level (stride) can be queried AFTER model creation via the `.feature_info` member
-* All models have a consistent pretrained weight loader that adapts last linear if necessary, and from 3 to 1 channel input if desired
-* High performance [reference training, validation, and inference scripts](https://rwightman.github.io/pytorch-image-models/scripts/) that work in several process/GPU modes:
-    * NVIDIA DDP w/ a single GPU per process, multiple processes with APEX present (AMP mixed-precision optional)
-    * PyTorch DistributedDataParallel w/ multi-gpu, single process (AMP disabled as it crashes when enabled)
-    * PyTorch w/ single GPU single process (AMP optional)
-* A dynamic global pool implementation that allows selecting from average pooling, max pooling, average + max, or concat([average, max]) at model creation. All global pooling is adaptive average by default and compatible with pretrained weights.
-* A 'Test Time Pool' wrapper that can wrap any of the included models and usually provides improved performance doing inference with input images larger than the training size. Idea adapted from original DPN implementation when I ported (https://github.com/cypw/DPNs)
-* Learning rate schedulers
-  * Ideas adopted from
-     * [AllenNLP schedulers](https://github.com/allenai/allennlp/tree/master/allennlp/training/learning_rate_schedulers)
-     * [FAIRseq lr_scheduler](https://github.com/pytorch/fairseq/tree/master/fairseq/optim/lr_scheduler)
-     * SGDR: Stochastic Gradient Descent with Warm Restarts (https://arxiv.org/abs/1608.03983)
-  * Schedulers include `step`, `cosine` w/ restarts, `tanh` w/ restarts, `plateau`
-* Optimizers:
-    * `rmsprop_tf` adapted from PyTorch RMSProp by myself. Reproduces much improved Tensorflow RMSProp behaviour.
-    * `radam` by [Liyuan Liu](https://github.com/LiyuanLucasLiu/RAdam) (https://arxiv.org/abs/1908.03265)
-    * `novograd` by [Masashi Kimura](https://github.com/convergence-lab/novograd) (https://arxiv.org/abs/1905.11286)
-    * `lookahead` adapted from impl by [Liam](https://github.com/alphadl/lookahead.pytorch) (https://arxiv.org/abs/1907.08610)
-    * `fused<name>` optimizers by name with [NVIDIA Apex](https://github.com/NVIDIA/apex/tree/master/apex/optimizers) installed
-    * `adamp` and `sgdp` by [Naver ClovAI](https://github.com/clovaai) (https://arxiv.org/abs/2006.08217)
-    * `adafactor` adapted from [FAIRSeq impl](https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py) (https://arxiv.org/abs/1804.04235)
-    * `adahessian` by [David Samuel](https://github.com/davda54/ada-hessian) (https://arxiv.org/abs/2006.00719)
-* Random Erasing from [Zhun Zhong](https://github.com/zhunzhong07/Random-Erasing/blob/master/transforms.py)  (https://arxiv.org/abs/1708.04896)
-* Mixup (https://arxiv.org/abs/1710.09412)
-* CutMix (https://arxiv.org/abs/1905.04899)
-* AutoAugment (https://arxiv.org/abs/1805.09501) and RandAugment (https://arxiv.org/abs/1909.13719) ImageNet configurations modeled after impl for EfficientNet training (https://github.com/tensorflow/tpu/blob/master/models/official/efficientnet/autoaugment.py)
-* AugMix w/ JSD loss (https://arxiv.org/abs/1912.02781), JSD w/ clean + augmented mixing support works with AutoAugment and RandAugment as well
-* SplitBachNorm - allows splitting batch norm layers between clean and augmented (auxiliary batch norm) data
-* DropPath aka "Stochastic Depth" (https://arxiv.org/abs/1603.09382) 
-* DropBlock (https://arxiv.org/abs/1810.12890)
-* Blur Pooling (https://arxiv.org/abs/1904.11486)
-* Space-to-Depth by [mrT23](https://github.com/mrT23/TResNet/blob/master/src/models/tresnet/layers/space_to_depth.py) (https://arxiv.org/abs/1801.04590) -- original paper?
-* Adaptive Gradient Clipping (https://arxiv.org/abs/2102.06171, https://github.com/deepmind/deepmind-research/tree/master/nfnets)
-* An extensive selection of channel and/or spatial attention modules:
-    * Bottleneck Transformer - https://arxiv.org/abs/2101.11605
-    * CBAM - https://arxiv.org/abs/1807.06521
-    * Effective Squeeze-Excitation (ESE) - https://arxiv.org/abs/1911.06667
-    * Efficient Channel Attention (ECA) - https://arxiv.org/abs/1910.03151
-    * Gather-Excite (GE) - https://arxiv.org/abs/1810.12348
-    * Global Context (GC) - https://arxiv.org/abs/1904.11492
-    * Halo - https://arxiv.org/abs/2103.12731
-    * Involution - https://arxiv.org/abs/2103.06255
-    * Lambda Layer - https://arxiv.org/abs/2102.08602
-    * Non-Local (NL) -  https://arxiv.org/abs/1711.07971
-    * Squeeze-and-Excitation (SE) - https://arxiv.org/abs/1709.01507
-    * Selective Kernel (SK) - (https://arxiv.org/abs/1903.06586
-    * Split (SPLAT) - https://arxiv.org/abs/2004.08955
-    * Shifted Window (SWIN) - https://arxiv.org/abs/2103.14030
-
-## Results
-
-Model validation results can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/results/) and in the [results tables](results/README.md)
-
-## Getting Started (Documentation)
-
-My current [documentation](https://rwightman.github.io/pytorch-image-models/) for `timm` covers the basics.
-
-[timmdocs](https://fastai.github.io/timmdocs/) is quickly becoming a much more comprehensive set of documentation for `timm`. A big thanks to [Aman Arora](https://github.com/amaarora) for his efforts creating timmdocs.
-
-[paperswithcode](https://paperswithcode.com/lib/timm) is a good resource for browsing the models within `timm`.
-
-## Train, Validation, Inference Scripts
-
-The root folder of the repository contains reference train, validation, and inference scripts that work with the included models and other features of this repository. They are adaptable for other datasets and use cases with a little hacking. See [documentation](https://rwightman.github.io/pytorch-image-models/scripts/) for some basics and [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) for some train examples that produce SOTA ImageNet results.
-
-## Awesome PyTorch Resources
-
-One of the greatest assets of PyTorch is the community and their contributions. A few of my favourite resources that pair well with the models and components here are listed below.
-
-### Object Detection, Instance and Semantic Segmentation
-* Detectron2 - https://github.com/facebookresearch/detectron2
-* Segmentation Models (Semantic) - https://github.com/qubvel/segmentation_models.pytorch
-* EfficientDet (Obj Det, Semantic soon) - https://github.com/rwightman/efficientdet-pytorch
-
-### Computer Vision / Image Augmentation
-* Albumentations - https://github.com/albumentations-team/albumentations
-* Kornia - https://github.com/kornia/kornia
-
-### Knowledge Distillation
-* RepDistiller - https://github.com/HobbitLong/RepDistiller
-* torchdistill - https://github.com/yoshitomo-matsubara/torchdistill
-
-### Metric Learning
-* PyTorch Metric Learning - https://github.com/KevinMusgrave/pytorch-metric-learning
-
-### Training / Frameworks
-* fastai - https://github.com/fastai/fastai
-
-## Licenses
-
-### Code
-The code here is licensed Apache 2.0. I've taken care to make sure any third party code included or adapted has compatible (permissive) licenses such as MIT, BSD, etc. I've made an effort to avoid any GPL / LGPL conflicts. That said, it is your responsibility to ensure you comply with licenses here and conditions of any dependent licenses. Where applicable, I've linked the sources/references for various components in docstrings. If you think I've missed anything please create an issue.
-
-### Pretrained Weights
-So far all of the pretrained weights available here are pretrained on ImageNet with a select few that have some additional pretraining (see extra note below). ImageNet was released for non-commercial research purposes only (http://www.image-net.org/download-faq). It's not clear what the implications of that are for the use of pretrained weights from that dataset. Any models I have trained with ImageNet are done for research purposes and one should assume that the original dataset license applies to the weights. It's best to seek legal advice if you intend to use the pretrained weights in a commercial product.
-
-#### Pretrained on more than ImageNet
-Several weights included or references here were pretrained with proprietary datasets that I do not have access to. These include the Facebook WSL, SSL, SWSL ResNe(Xt) and the Google Noisy Student EfficientNet models. The Facebook models have an explicit non-commercial license (CC-BY-NC 4.0, https://github.com/facebookresearch/semi-supervised-ImageNet1K-models, https://github.com/facebookresearch/WSL-Images). The Google models do not appear to have any restriction beyond the Apache 2.0 license (and ImageNet concerns). In either case, you should contact Facebook or Google with any questions.
-
-## Citing
-
-### BibTeX
-
-```
-@misc{rw2019timm,
-  author = {Ross Wightman},
-  title = {PyTorch Image Models},
-  year = {2019},
-  publisher = {GitHub},
-  journal = {GitHub repository},
-  doi = {10.5281/zenodo.4414861},
-  howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
-}
-```
-
-### Latest DOI
-
-[![DOI](https://zenodo.org/badge/168799526.svg)](https://zenodo.org/badge/latestdoi/168799526)
+        /home/Gluon_ResNet50_v1c_for_PyTorch/test/output/0
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README_raw.md b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README_raw.md
new file mode 100644
index 0000000000..76261cccf2
--- /dev/null
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/README_raw.md
@@ -0,0 +1,399 @@
+# PyTorch Image Models
+- [Sponsors](#sponsors)
+- [What's New](#whats-new)
+- [Introduction](#introduction)
+- [Models](#models)
+- [Features](#features)
+- [Results](#results)
+- [Getting Started (Documentation)](#getting-started-documentation)
+- [Train, Validation, Inference Scripts](#train-validation-inference-scripts)
+- [Awesome PyTorch Resources](#awesome-pytorch-resources)
+- [Licenses](#licenses)
+- [Citing](#citing)
+
+## Sponsors
+
+A big thank you to my [GitHub Sponsors](https://github.com/sponsors/rwightman) for their support!
+
+In addition to the sponsors at the link above, I've received hardware and/or cloud resources from
+* Nvidia (https://www.nvidia.com/en-us/)
+* TFRC (https://www.tensorflow.org/tfrc)
+
+I'm fortunate to be able to dedicate significant time and money of my own supporting this and other open source projects. However, as the projects increase in scope, outside support is needed to continue with the current trajectory of hardware, infrastructure, and electricty costs.
+
+## What's New
+
+### June 20, 2021
+* Release Vision Transformer 'AugReg' weights from [How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers](https://arxiv.org/abs/2106.10270)
+  * .npz weight loading support added, can load any of the 50K+ weights from the [AugReg series](https://console.cloud.google.com/storage/browser/vit_models/augreg)
+  * See [example notebook](https://colab.research.google.com/github/google-research/vision_transformer/blob/master/vit_jax_augreg.ipynb) from official impl for navigating the augreg weights
+  * Replaced all default weights w/ best AugReg variant (if possible). All AugReg 21k classifiers work.
+    * Highlights: `vit_large_patch16_384` (87.1 top-1), `vit_large_r50_s32_384` (86.2 top-1), `vit_base_patch16_384` (86.0 top-1)
+  * `vit_deit_*` renamed to just `deit_*`
+  * Remove my old small model, replace with DeiT compatible small w/ AugReg weights
+* Add 1st training of my `gmixer_24_224` MLP /w GLU, 78.1 top-1 w/ 25M params.
+* Add weights from official ResMLP release (https://github.com/facebookresearch/deit)
+* Add `eca_nfnet_l2` weights from my 'lightweight' series. 84.7 top-1 at 384x384.
+* Add distilled BiT 50x1 student and 152x2 Teacher weights from  [Knowledge distillation: A good teacher is patient and consistent](https://arxiv.org/abs/2106.05237)
+* NFNets and ResNetV2-BiT models work w/ Pytorch XLA now
+  * weight standardization uses F.batch_norm instead of std_mean (std_mean wasn't lowered)
+  * eps values adjusted, will be slight differences but should be quite close
+* Improve test coverage and classifier interface of non-conv (vision transformer and mlp) models
+* Cleanup a few classifier / flatten details for models w/ conv classifiers or early global pool
+* Please report any regressions, this PR touched quite a few models.
+
+### June 8, 2021
+* Add first ResMLP weights, trained in PyTorch XLA on TPU-VM w/ my XLA branch. 24 block variant, 79.2 top-1.
+* Add ResNet51-Q model w/ pretrained weights at 82.36 top-1.
+  * NFNet inspired block layout with quad layer stem and no maxpool
+  * Same param count (35.7M) and throughput as ResNetRS-50 but +1.5 top-1 @ 224x224 and +2.5 top-1 at 288x288
+
+### May 25, 2021
+* Add LeViT, Visformer, ConViT (PR by Aman Arora), Twins (PR by paper authors) transformer models
+* Add ResMLP and gMLP MLP vision models to the existing MLP Mixer impl
+* Fix a number of torchscript issues with various vision transformer models
+* Cleanup input_size/img_size override handling and improve testing / test coverage for all vision transformer and MLP models
+* More flexible pos embedding resize (non-square) for ViT and TnT. Thanks [Alexander Soare](https://github.com/alexander-soare)
+* Add `efficientnetv2_rw_m` model and weights (started training before official code). 84.8 top-1, 53M params.
+
+### May 14, 2021
+* Add EfficientNet-V2 official model defs w/ ported weights from official [Tensorflow/Keras](https://github.com/google/automl/tree/master/efficientnetv2) impl.
+  * 1k trained variants: `tf_efficientnetv2_s/m/l`
+  * 21k trained variants: `tf_efficientnetv2_s/m/l_in21k`
+  * 21k pretrained -> 1k fine-tuned: `tf_efficientnetv2_s/m/l_in21ft1k`
+  * v2 models w/ v1 scaling: `tf_efficientnetv2_b0` through `b3`
+  * Rename my prev V2 guess `efficientnet_v2s` -> `efficientnetv2_rw_s`
+  * Some blank `efficientnetv2_*` models in-place for future native PyTorch training
+
+### May 5, 2021
+* Add MLP-Mixer models and port pretrained weights from [Google JAX impl](https://github.com/google-research/vision_transformer/tree/linen)
+* Add CaiT models and pretrained weights from [FB](https://github.com/facebookresearch/deit)
+* Add ResNet-RS models and weights from [TF](https://github.com/tensorflow/tpu/tree/master/models/official/resnet/resnet_rs). Thanks [Aman Arora](https://github.com/amaarora)
+* Add CoaT models and weights. Thanks [Mohammed Rizin](https://github.com/morizin)
+* Add new ImageNet-21k weights & finetuned weights for TResNet, MobileNet-V3, ViT models. Thanks [mrT](https://github.com/mrT23)
+* Add GhostNet models and weights. Thanks [Kai Han](https://github.com/iamhankai)
+* Update ByoaNet attention modules
+   * Improve SA module inits
+   * Hack together experimental stand-alone Swin based attn module and `swinnet`
+   * Consistent '26t' model defs for experiments.
+* Add improved Efficientnet-V2S (prelim model def) weights. 83.8 top-1.
+* WandB logging support
+
+### April 13, 2021
+* Add Swin Transformer models and weights from https://github.com/microsoft/Swin-Transformer
+
+### April 12, 2021
+* Add ECA-NFNet-L1 (slimmed down F1 w/ SiLU, 41M params) trained with this code. 84% top-1 @ 320x320. Trained at 256x256.
+* Add EfficientNet-V2S model (unverified model definition) weights. 83.3 top-1 @ 288x288. Only trained single res 224. Working on progressive training.
+* Add ByoaNet model definition (Bring-your-own-attention) w/ SelfAttention block and corresponding SA/SA-like modules and model defs
+  * Lambda Networks - https://arxiv.org/abs/2102.08602
+  * Bottleneck Transformers - https://arxiv.org/abs/2101.11605
+  * Halo Nets - https://arxiv.org/abs/2103.12731
+* Adabelief optimizer contributed by Juntang Zhuang
+
+### April 1, 2021
+* Add snazzy `benchmark.py` script for bulk `timm` model benchmarking of train and/or inference
+* Add Pooling-based Vision Transformer (PiT) models (from https://github.com/naver-ai/pit)
+  * Merged distilled variant into main for torchscript compatibility
+  * Some `timm` cleanup/style tweaks and weights have hub download support
+* Cleanup Vision Transformer (ViT) models
+  * Merge distilled (DeiT) model into main so that torchscript can work
+  * Support updated weight init (defaults to old still) that closer matches original JAX impl (possibly better training from scratch)
+  * Separate hybrid model defs into different file and add several new model defs to fiddle with, support patch_size != 1 for hybrids
+  * Fix fine-tuning num_class changes (PiT and ViT) and pos_embed resizing (Vit) with distilled variants
+  * nn.Sequential for block stack (does not break downstream compat)
+* TnT (Transformer-in-Transformer) models contributed by author (from https://gitee.com/mindspore/mindspore/tree/master/model_zoo/research/cv/TNT)
+* Add RegNetY-160 weights from DeiT teacher model
+* Add new NFNet-L0 w/ SE attn (rename `nfnet_l0b`->`nfnet_l0`) weights 82.75 top-1 @ 288x288
+* Some fixes/improvements for TFDS dataset wrapper
+
+### March 17, 2021
+* Add new ECA-NFNet-L0 (rename `nfnet_l0c`->`eca_nfnet_l0`) weights trained by myself.
+  * 82.6 top-1 @ 288x288, 82.8 @ 320x320, trained at 224x224
+  * Uses SiLU activation, approx 2x faster than `dm_nfnet_f0` and 50% faster than `nfnet_f0s` w/ 1/3 param count
+* Integrate [Hugging Face model hub](https://huggingface.co/models) into timm create_model and default_cfg handling for pretrained weight and config sharing (more on this soon!)
+* Merge HardCoRe NAS models contributed by https://github.com/yoniaflalo
+* Merge PyTorch trained EfficientNet-EL and pruned ES/EL variants contributed by [DeGirum](https://github.com/DeGirum)
+
+
+### March 7, 2021
+* First 0.4.x PyPi release w/ NFNets (& related), ByoB (GPU-Efficient, RepVGG, etc).
+* Change feature extraction for pre-activation nets (NFNets, ResNetV2) to return features before activation.
+* Tested with PyTorch 1.8 release. Updated CI to use 1.8.
+* Benchmarked several arch on RTX 3090, Titan RTX, and V100 across 1.7.1, 1.8, NGC 20.12, and 21.02. Some interesting performance variations to take note of https://gist.github.com/rwightman/bb59f9e245162cee0e38bd66bd8cd77f
+
+### Feb 18, 2021
+* Add pretrained weights and model variants for NFNet-F* models from [DeepMind Haiku impl](https://github.com/deepmind/deepmind-research/tree/master/nfnets).
+  * Models are prefixed with `dm_`. They require SAME padding conv, skipinit enabled, and activation gains applied in act fn.
+  * These models are big, expect to run out of GPU memory. With the GELU activiation + other options, they are roughly 1/2 the inference speed of my SiLU PyTorch optimized `s` variants.
+  * Original model results are based on pre-processing that is not the same as all other models so you'll see different results in the results csv (once updated).
+  * Matching the original pre-processing as closely as possible I get these results:
+    * `dm_nfnet_f6` - 86.352
+    * `dm_nfnet_f5` - 86.100
+    * `dm_nfnet_f4` - 85.834
+    * `dm_nfnet_f3` - 85.676
+    * `dm_nfnet_f2` - 85.178
+    * `dm_nfnet_f1` - 84.696
+    * `dm_nfnet_f0` - 83.464
+
+### Feb 16, 2021
+* Add Adaptive Gradient Clipping (AGC) as per https://arxiv.org/abs/2102.06171. Integrated w/ PyTorch gradient clipping via mode arg that defaults to prev 'norm' mode. For backward arg compat, clip-grad arg must be specified to enable when using train.py.
+  * AGC w/ default clipping factor `--clip-grad .01 --clip-mode agc`
+  * PyTorch global norm of 1.0 (old behaviour, always norm), `--clip-grad 1.0`
+  * PyTorch value clipping of 10, `--clip-grad 10. --clip-mode value`
+  * AGC performance is definitely sensitive to the clipping factor. More experimentation needed to determine good values for smaller batch sizes and optimizers besides those in paper. So far I've found .001-.005 is necessary for stable RMSProp training w/ NFNet/NF-ResNet.
+
+### Feb 12, 2021
+* Update Normalization-Free nets to include new NFNet-F (https://arxiv.org/abs/2102.06171) model defs
+
+### Feb 10, 2021
+* First Normalization-Free model training experiments done,
+  * nf_resnet50 - 80.68 top-1 @ 288x288, 80.31 @ 256x256
+  * nf_regnet_b1 - 79.30 @ 288x288, 78.75 @ 256x256
+* More model archs, incl a flexible ByobNet backbone ('Bring-your-own-blocks')
+  * GPU-Efficient-Networks (https://github.com/idstcv/GPU-Efficient-Networks), impl in `byobnet.py`
+  * RepVGG (https://github.com/DingXiaoH/RepVGG), impl in `byobnet.py`
+  * classic VGG (from torchvision, impl in `vgg.py`)
+* Refinements to normalizer layer arg handling and normalizer+act layer handling in some models
+* Default AMP mode changed to native PyTorch AMP instead of APEX. Issues not being fixed with APEX. Native works with `--channels-last` and `--torchscript` model training, APEX does not.
+* Fix a few bugs introduced since last pypi release
+
+### Feb 8, 2021
+* Add several ResNet weights with ECA attention. 26t & 50t trained @ 256, test @ 320. 269d train @ 256, fine-tune @320, test @ 352.
+  * `ecaresnet26t` - 79.88 top-1 @ 320x320, 79.08 @ 256x256
+  * `ecaresnet50t` - 82.35 top-1 @ 320x320, 81.52 @ 256x256
+  * `ecaresnet269d` - 84.93 top-1 @ 352x352, 84.87 @ 320x320
+* Remove separate tiered (`t`) vs tiered_narrow (`tn`) ResNet model defs, all `tn` changed to `t` and `t` models removed (`seresnext26t_32x4d` only model w/ weights that was removed).
+* Support model default_cfgs with separate train vs test resolution `test_input_size` and remove extra `_320` suffix ResNet model defs that were just for test.
+
+### Jan 30, 2021
+* Add initial "Normalization Free" NF-RegNet-B* and NF-ResNet model definitions based on [paper](https://arxiv.org/abs/2101.08692)
+
+### Jan 25, 2021
+* Add ResNetV2 Big Transfer (BiT) models w/ ImageNet-1k and 21k weights from https://github.com/google-research/big_transfer
+* Add official R50+ViT-B/16 hybrid models + weights from https://github.com/google-research/vision_transformer
+* ImageNet-21k ViT weights are added w/ model defs and representation layer (pre logits) support
+  * NOTE: ImageNet-21k classifier heads were zero'd in original weights, they are only useful for transfer learning
+* Add model defs and weights for DeiT Vision Transformer models from https://github.com/facebookresearch/deit
+* Refactor dataset classes into ImageDataset/IterableImageDataset + dataset specific parser classes
+* Add Tensorflow-Datasets (TFDS) wrapper to allow use of TFDS image classification sets with train script
+  * Ex: `train.py /data/tfds --dataset tfds/oxford_iiit_pet --val-split test --model resnet50 -b 256 --amp --num-classes 37 --opt adamw --lr 3e-4 --weight-decay .001 --pretrained -j 2`
+* Add improved .tar dataset parser that reads images from .tar, folder of .tar files, or .tar within .tar
+  * Run validation on full ImageNet-21k directly from tar w/ BiT model: `validate.py /data/fall11_whole.tar --model resnetv2_50x1_bitm_in21k --amp`
+* Models in this update should be stable w/ possible exception of ViT/BiT, possibility of some regressions with train/val scripts and dataset handling
+
+### Jan 3, 2021
+* Add SE-ResNet-152D weights
+  * 256x256 val, 0.94 crop top-1 - 83.75
+  * 320x320 val, 1.0 crop - 84.36
+* Update [results files](results/)
+
+
+## Introduction
+
+Py**T**orch **Im**age **M**odels (`timm`) is a collection of image models, layers, utilities, optimizers, schedulers, data-loaders / augmentations, and reference training / validation scripts that aim to pull together a wide variety of SOTA models with ability to reproduce ImageNet training results.
+
+The work of many others is present here. I've tried to make sure all source material is acknowledged via links to github, arxiv papers, etc in the README, documentation, and code docstrings. Please let me know if I missed anything.
+
+## Models
+
+All model architecture families include variants with pretrained weights. There are specific model variants without any weights, it is NOT a bug. Help training new or better weights is always appreciated. Here are some example [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) to get you started.
+
+A full version of the list below with source links can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/models/).
+
+* Big Transfer ResNetV2 (BiT) - https://arxiv.org/abs/1912.11370
+* Bottleneck Transformers - https://arxiv.org/abs/2101.11605
+* CaiT (Class-Attention in Image Transformers) - https://arxiv.org/abs/2103.17239
+* CoaT (Co-Scale Conv-Attentional Image Transformers) - https://arxiv.org/abs/2104.06399
+* ConViT (Soft Convolutional Inductive Biases Vision Transformers)- https://arxiv.org/abs/2103.10697
+* CspNet (Cross-Stage Partial Networks) - https://arxiv.org/abs/1911.11929
+* DeiT (Vision Transformer) - https://arxiv.org/abs/2012.12877
+* DenseNet - https://arxiv.org/abs/1608.06993
+* DLA - https://arxiv.org/abs/1707.06484
+* DPN (Dual-Path Network) - https://arxiv.org/abs/1707.01629
+* EfficientNet (MBConvNet Family)
+    * EfficientNet NoisyStudent (B0-B7, L2) - https://arxiv.org/abs/1911.04252
+    * EfficientNet AdvProp (B0-B8) - https://arxiv.org/abs/1911.09665
+    * EfficientNet (B0-B7) - https://arxiv.org/abs/1905.11946
+    * EfficientNet-EdgeTPU (S, M, L) - https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html
+    * EfficientNet V2 - https://arxiv.org/abs/2104.00298
+    * FBNet-C - https://arxiv.org/abs/1812.03443
+    * MixNet - https://arxiv.org/abs/1907.09595
+    * MNASNet B1, A1 (Squeeze-Excite), and Small - https://arxiv.org/abs/1807.11626
+    * MobileNet-V2 - https://arxiv.org/abs/1801.04381
+    * Single-Path NAS - https://arxiv.org/abs/1904.02877
+* GhostNet - https://arxiv.org/abs/1911.11907
+* gMLP - https://arxiv.org/abs/2105.08050
+* GPU-Efficient Networks - https://arxiv.org/abs/2006.14090
+* Halo Nets - https://arxiv.org/abs/2103.12731
+* HardCoRe-NAS - https://arxiv.org/abs/2102.11646
+* HRNet - https://arxiv.org/abs/1908.07919
+* Inception-V3 - https://arxiv.org/abs/1512.00567
+* Inception-ResNet-V2 and Inception-V4 - https://arxiv.org/abs/1602.07261
+* Lambda Networks - https://arxiv.org/abs/2102.08602
+* LeViT (Vision Transformer in ConvNet's Clothing) - https://arxiv.org/abs/2104.01136
+* MLP-Mixer - https://arxiv.org/abs/2105.01601
+* MobileNet-V3 (MBConvNet w/ Efficient Head) - https://arxiv.org/abs/1905.02244
+* NASNet-A - https://arxiv.org/abs/1707.07012
+* NFNet-F - https://arxiv.org/abs/2102.06171
+* NF-RegNet / NF-ResNet - https://arxiv.org/abs/2101.08692
+* PNasNet - https://arxiv.org/abs/1712.00559
+* Pooling-based Vision Transformer (PiT) - https://arxiv.org/abs/2103.16302
+* RegNet - https://arxiv.org/abs/2003.13678
+* RepVGG - https://arxiv.org/abs/2101.03697
+* ResMLP - https://arxiv.org/abs/2105.03404
+* ResNet/ResNeXt
+    * ResNet (v1b/v1.5) - https://arxiv.org/abs/1512.03385
+    * ResNeXt - https://arxiv.org/abs/1611.05431
+    * 'Bag of Tricks' / Gluon C, D, E, S variations - https://arxiv.org/abs/1812.01187
+    * Weakly-supervised (WSL) Instagram pretrained / ImageNet tuned ResNeXt101 - https://arxiv.org/abs/1805.00932
+    * Semi-supervised (SSL) / Semi-weakly Supervised (SWSL) ResNet/ResNeXts - https://arxiv.org/abs/1905.00546
+    * ECA-Net (ECAResNet) - https://arxiv.org/abs/1910.03151v4
+    * Squeeze-and-Excitation Networks (SEResNet) - https://arxiv.org/abs/1709.01507
+    * ResNet-RS - https://arxiv.org/abs/2103.07579
+* Res2Net - https://arxiv.org/abs/1904.01169
+* ResNeSt - https://arxiv.org/abs/2004.08955
+* ReXNet - https://arxiv.org/abs/2007.00992
+* SelecSLS - https://arxiv.org/abs/1907.00837
+* Selective Kernel Networks - https://arxiv.org/abs/1903.06586
+* Swin Transformer - https://arxiv.org/abs/2103.14030
+* Transformer-iN-Transformer (TNT) - https://arxiv.org/abs/2103.00112
+* TResNet - https://arxiv.org/abs/2003.13630
+* Twins (Spatial Attention in Vision Transformers) - https://arxiv.org/pdf/2104.13840.pdf
+* Vision Transformer - https://arxiv.org/abs/2010.11929
+* VovNet V2 and V1 - https://arxiv.org/abs/1911.06667
+* Xception - https://arxiv.org/abs/1610.02357
+* Xception (Modified Aligned, Gluon) - https://arxiv.org/abs/1802.02611
+* Xception (Modified Aligned, TF) - https://arxiv.org/abs/1802.02611
+
+## Features
+
+Several (less common) features that I often utilize in my projects are included. Many of their additions are the reason why I maintain my own set of models, instead of using others' via PIP:
+
+* All models have a common default configuration interface and API for
+    * accessing/changing the classifier - `get_classifier` and `reset_classifier`
+    * doing a forward pass on just the features - `forward_features` (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
+    * these makes it easy to write consistent network wrappers that work with any of the models
+* All models support multi-scale feature map extraction (feature pyramids) via create_model (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
+    * `create_model(name, features_only=True, out_indices=..., output_stride=...)`
+    * `out_indices` creation arg specifies which feature maps to return, these indices are 0 based and generally correspond to the `C(i + 1)` feature level.
+    * `output_stride` creation arg controls output stride of the network by using dilated convolutions. Most networks are stride 32 by default. Not all networks support this.
+    * feature map channel counts, reduction level (stride) can be queried AFTER model creation via the `.feature_info` member
+* All models have a consistent pretrained weight loader that adapts last linear if necessary, and from 3 to 1 channel input if desired
+* High performance [reference training, validation, and inference scripts](https://rwightman.github.io/pytorch-image-models/scripts/) that work in several process/GPU modes:
+    * NVIDIA DDP w/ a single GPU per process, multiple processes with APEX present (AMP mixed-precision optional)
+    * PyTorch DistributedDataParallel w/ multi-gpu, single process (AMP disabled as it crashes when enabled)
+    * PyTorch w/ single GPU single process (AMP optional)
+* A dynamic global pool implementation that allows selecting from average pooling, max pooling, average + max, or concat([average, max]) at model creation. All global pooling is adaptive average by default and compatible with pretrained weights.
+* A 'Test Time Pool' wrapper that can wrap any of the included models and usually provides improved performance doing inference with input images larger than the training size. Idea adapted from original DPN implementation when I ported (https://github.com/cypw/DPNs)
+* Learning rate schedulers
+  * Ideas adopted from
+     * [AllenNLP schedulers](https://github.com/allenai/allennlp/tree/master/allennlp/training/learning_rate_schedulers)
+     * [FAIRseq lr_scheduler](https://github.com/pytorch/fairseq/tree/master/fairseq/optim/lr_scheduler)
+     * SGDR: Stochastic Gradient Descent with Warm Restarts (https://arxiv.org/abs/1608.03983)
+  * Schedulers include `step`, `cosine` w/ restarts, `tanh` w/ restarts, `plateau`
+* Optimizers:
+    * `rmsprop_tf` adapted from PyTorch RMSProp by myself. Reproduces much improved Tensorflow RMSProp behaviour.
+    * `radam` by [Liyuan Liu](https://github.com/LiyuanLucasLiu/RAdam) (https://arxiv.org/abs/1908.03265)
+    * `novograd` by [Masashi Kimura](https://github.com/convergence-lab/novograd) (https://arxiv.org/abs/1905.11286)
+    * `lookahead` adapted from impl by [Liam](https://github.com/alphadl/lookahead.pytorch) (https://arxiv.org/abs/1907.08610)
+    * `fused<name>` optimizers by name with [NVIDIA Apex](https://github.com/NVIDIA/apex/tree/master/apex/optimizers) installed
+    * `adamp` and `sgdp` by [Naver ClovAI](https://github.com/clovaai) (https://arxiv.org/abs/2006.08217)
+    * `adafactor` adapted from [FAIRSeq impl](https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py) (https://arxiv.org/abs/1804.04235)
+    * `adahessian` by [David Samuel](https://github.com/davda54/ada-hessian) (https://arxiv.org/abs/2006.00719)
+* Random Erasing from [Zhun Zhong](https://github.com/zhunzhong07/Random-Erasing/blob/master/transforms.py)  (https://arxiv.org/abs/1708.04896)
+* Mixup (https://arxiv.org/abs/1710.09412)
+* CutMix (https://arxiv.org/abs/1905.04899)
+* AutoAugment (https://arxiv.org/abs/1805.09501) and RandAugment (https://arxiv.org/abs/1909.13719) ImageNet configurations modeled after impl for EfficientNet training (https://github.com/tensorflow/tpu/blob/master/models/official/efficientnet/autoaugment.py)
+* AugMix w/ JSD loss (https://arxiv.org/abs/1912.02781), JSD w/ clean + augmented mixing support works with AutoAugment and RandAugment as well
+* SplitBachNorm - allows splitting batch norm layers between clean and augmented (auxiliary batch norm) data
+* DropPath aka "Stochastic Depth" (https://arxiv.org/abs/1603.09382) 
+* DropBlock (https://arxiv.org/abs/1810.12890)
+* Blur Pooling (https://arxiv.org/abs/1904.11486)
+* Space-to-Depth by [mrT23](https://github.com/mrT23/TResNet/blob/master/src/models/tresnet/layers/space_to_depth.py) (https://arxiv.org/abs/1801.04590) -- original paper?
+* Adaptive Gradient Clipping (https://arxiv.org/abs/2102.06171, https://github.com/deepmind/deepmind-research/tree/master/nfnets)
+* An extensive selection of channel and/or spatial attention modules:
+    * Bottleneck Transformer - https://arxiv.org/abs/2101.11605
+    * CBAM - https://arxiv.org/abs/1807.06521
+    * Effective Squeeze-Excitation (ESE) - https://arxiv.org/abs/1911.06667
+    * Efficient Channel Attention (ECA) - https://arxiv.org/abs/1910.03151
+    * Gather-Excite (GE) - https://arxiv.org/abs/1810.12348
+    * Global Context (GC) - https://arxiv.org/abs/1904.11492
+    * Halo - https://arxiv.org/abs/2103.12731
+    * Involution - https://arxiv.org/abs/2103.06255
+    * Lambda Layer - https://arxiv.org/abs/2102.08602
+    * Non-Local (NL) -  https://arxiv.org/abs/1711.07971
+    * Squeeze-and-Excitation (SE) - https://arxiv.org/abs/1709.01507
+    * Selective Kernel (SK) - (https://arxiv.org/abs/1903.06586
+    * Split (SPLAT) - https://arxiv.org/abs/2004.08955
+    * Shifted Window (SWIN) - https://arxiv.org/abs/2103.14030
+
+## Results
+
+Model validation results can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/results/) and in the [results tables](results/README.md)
+
+## Getting Started (Documentation)
+
+My current [documentation](https://rwightman.github.io/pytorch-image-models/) for `timm` covers the basics.
+
+[timmdocs](https://fastai.github.io/timmdocs/) is quickly becoming a much more comprehensive set of documentation for `timm`. A big thanks to [Aman Arora](https://github.com/amaarora) for his efforts creating timmdocs.
+
+[paperswithcode](https://paperswithcode.com/lib/timm) is a good resource for browsing the models within `timm`.
+
+## Train, Validation, Inference Scripts
+
+The root folder of the repository contains reference train, validation, and inference scripts that work with the included models and other features of this repository. They are adaptable for other datasets and use cases with a little hacking. See [documentation](https://rwightman.github.io/pytorch-image-models/scripts/) for some basics and [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) for some train examples that produce SOTA ImageNet results.
+
+## Awesome PyTorch Resources
+
+One of the greatest assets of PyTorch is the community and their contributions. A few of my favourite resources that pair well with the models and components here are listed below.
+
+### Object Detection, Instance and Semantic Segmentation
+* Detectron2 - https://github.com/facebookresearch/detectron2
+* Segmentation Models (Semantic) - https://github.com/qubvel/segmentation_models.pytorch
+* EfficientDet (Obj Det, Semantic soon) - https://github.com/rwightman/efficientdet-pytorch
+
+### Computer Vision / Image Augmentation
+* Albumentations - https://github.com/albumentations-team/albumentations
+* Kornia - https://github.com/kornia/kornia
+
+### Knowledge Distillation
+* RepDistiller - https://github.com/HobbitLong/RepDistiller
+* torchdistill - https://github.com/yoshitomo-matsubara/torchdistill
+
+### Metric Learning
+* PyTorch Metric Learning - https://github.com/KevinMusgrave/pytorch-metric-learning
+
+### Training / Frameworks
+* fastai - https://github.com/fastai/fastai
+
+## Licenses
+
+### Code
+The code here is licensed Apache 2.0. I've taken care to make sure any third party code included or adapted has compatible (permissive) licenses such as MIT, BSD, etc. I've made an effort to avoid any GPL / LGPL conflicts. That said, it is your responsibility to ensure you comply with licenses here and conditions of any dependent licenses. Where applicable, I've linked the sources/references for various components in docstrings. If you think I've missed anything please create an issue.
+
+### Pretrained Weights
+So far all of the pretrained weights available here are pretrained on ImageNet with a select few that have some additional pretraining (see extra note below). ImageNet was released for non-commercial research purposes only (http://www.image-net.org/download-faq). It's not clear what the implications of that are for the use of pretrained weights from that dataset. Any models I have trained with ImageNet are done for research purposes and one should assume that the original dataset license applies to the weights. It's best to seek legal advice if you intend to use the pretrained weights in a commercial product.
+
+#### Pretrained on more than ImageNet
+Several weights included or references here were pretrained with proprietary datasets that I do not have access to. These include the Facebook WSL, SSL, SWSL ResNe(Xt) and the Google Noisy Student EfficientNet models. The Facebook models have an explicit non-commercial license (CC-BY-NC 4.0, https://github.com/facebookresearch/semi-supervised-ImageNet1K-models, https://github.com/facebookresearch/WSL-Images). The Google models do not appear to have any restriction beyond the Apache 2.0 license (and ImageNet concerns). In either case, you should contact Facebook or Google with any questions.
+
+## Citing
+
+### BibTeX
+
+```
+@misc{rw2019timm,
+  author = {Ross Wightman},
+  title = {PyTorch Image Models},
+  year = {2019},
+  publisher = {GitHub},
+  journal = {GitHub repository},
+  doi = {10.5281/zenodo.4414861},
+  howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
+}
+```
+
+### Latest DOI
+
+[![DOI](https://zenodo.org/badge/168799526.svg)](https://zenodo.org/badge/latestdoi/168799526)
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_1p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_1p.sh
index 4270119d12..0985f163ee 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_1p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_1p.sh
@@ -74,10 +74,16 @@ fi
 #################启动训练脚本#################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
 
-python3 train.py \
+# 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
+
+
+nohup python3.7 train.py \
     ${data_path} \
     -b ${batch_size} \
     --model gluon_resnet50_v1c \
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_8p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_8p.sh
index f52176cfcd..bf6c5f742d 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_8p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_full_8p.sh
@@ -63,12 +63,18 @@ fi
 #################启动训练脚本#################
 # 训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
+
+# 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
+
 
 for i in $(seq 0 7)
 do
-    python3 train.py \
+    nohup python3.7 train.py \
         ${data_path} \
         -b ${batch_size} \
         --model gluon_resnet50_v1c \
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_1p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_1p.sh
index dcdf1e17e4..04c2d333b4 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_1p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_1p.sh
@@ -47,8 +47,6 @@ else
     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`
@@ -61,7 +59,6 @@ else
     test_path_dir=${cur_path}/test
 fi
 
-
 #################创建日志输出目录，不需要修改#################
 if [ -d ${test_path_dir}/output/${ASCEND_DEVICE_ID} ];then
     rm -rf ${test_path_dir}/output/${ASCEND_DEVICE_ID}
@@ -74,10 +71,15 @@ fi
 #################启动训练脚本#################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
 
-python3 train.py \
+# 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
+
+nohup python3.7 train.py \
     ${data_path} \
     -b ${batch_size} \
     --model gluon_resnet50_v1c \
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_8p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_8p.sh
index 1d634eed55..c32a993a64 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_8p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1c_for_PyTorch/test/train_performance_8p.sh
@@ -63,12 +63,17 @@ fi
 #################启动训练脚本#################
 # 训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
+# 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
+
 
 for i in $(seq 0 7)
 do
-    python3 train.py \
+    nohup python3.7 train.py \
         ${data_path} \
         -b ${batch_size} \
         --model gluon_resnet50_v1c \
-- 
Gitee


From 0e918454975c0360c0bec2ca05451b27097d041d Mon Sep 17 00:00:00 2001
From: Bailang <bailang12@h-partners.com>
Date: Mon, 9 May 2022 15:11:42 +0800
Subject: [PATCH 5/5] =?UTF-8?q?[=E8=87=AA=E7=A0=94][PyTorch]Gluon=5FResNet?=
 =?UTF-8?q?50=5Fv1d=5Ffor=5FPyTorch=E6=A8=A1=E5=9E=8B=E8=AE=AD=E7=BB=83?=
 =?UTF-8?q?=E5=90=AF=E5=8A=A8=E8=84=9A=E6=9C=AC=E5=8F=AA=E7=95=99test?=
 =?UTF-8?q?=E4=B8=8Bshell=E8=84=9A=E6=9C=AC=EF=BC=8C=E4=BC=98=E5=8C=96test?=
 =?UTF-8?q?=E4=B8=8B=E8=84=9A=E6=9C=AC?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Signed-off-by: Bailang <bailang12@h-partners.com>
---
 .../Gluon_ResNet50_v1d_for_PyTorch/README.md  | 414 ++----------------
 .../README_raw.md                             | 399 +++++++++++++++++
 .../test/train_full_1p.sh                     |  12 +-
 .../test/train_full_8p.sh                     |  12 +-
 .../test/train_performance_1p.sh              |  12 +-
 .../test/train_performance_8p.sh              |  12 +-
 6 files changed, 461 insertions(+), 400 deletions(-)
 create mode 100644 PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README_raw.md

diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README.md b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README.md
index 76261cccf2..c3df4f61b9 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README.md
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README.md
@@ -1,399 +1,37 @@
-# PyTorch Image Models
-- [Sponsors](#sponsors)
-- [What's New](#whats-new)
-- [Introduction](#introduction)
-- [Models](#models)
-- [Features](#features)
-- [Results](#results)
-- [Getting Started (Documentation)](#getting-started-documentation)
-- [Train, Validation, Inference Scripts](#train-validation-inference-scripts)
-- [Awesome PyTorch Resources](#awesome-pytorch-resources)
-- [Licenses](#licenses)
-- [Citing](#citing)
+一、训练流程：
+    
 
-## Sponsors
+单卡训练流程：
 
-A big thank you to my [GitHub Sponsors](https://github.com/sponsors/rwightman) for their support!
+    1.安装环境  
+    2.开始训练
+        bash ./test/train_full_1p.sh  --data_path=数据集路径            # 精度训练
+        bash ./test/train_performance_1p.sh  --data_path=数据集路径    # 性能训练
 
-In addition to the sponsors at the link above, I've received hardware and/or cloud resources from
-* Nvidia (https://www.nvidia.com/en-us/)
-* TFRC (https://www.tensorflow.org/tfrc)
+ 
+多卡训练流程
 
-I'm fortunate to be able to dedicate significant time and money of my own supporting this and other open source projects. However, as the projects increase in scope, outside support is needed to continue with the current trajectory of hardware, infrastructure, and electricty costs.
+    1.安装环境
+    2.开始训练
+        bash ./test/train_full_8p.sh  --data_path=数据集路径          # 精度训练
+        bash ./test/train_performance_8p.sh  --data_path=数据集路径     # 性能训练
 
-## What's New
 
-### June 20, 2021
-* Release Vision Transformer 'AugReg' weights from [How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers](https://arxiv.org/abs/2106.10270)
-  * .npz weight loading support added, can load any of the 50K+ weights from the [AugReg series](https://console.cloud.google.com/storage/browser/vit_models/augreg)
-  * See [example notebook](https://colab.research.google.com/github/google-research/vision_transformer/blob/master/vit_jax_augreg.ipynb) from official impl for navigating the augreg weights
-  * Replaced all default weights w/ best AugReg variant (if possible). All AugReg 21k classifiers work.
-    * Highlights: `vit_large_patch16_384` (87.1 top-1), `vit_large_r50_s32_384` (86.2 top-1), `vit_base_patch16_384` (86.0 top-1)
-  * `vit_deit_*` renamed to just `deit_*`
-  * Remove my old small model, replace with DeiT compatible small w/ AugReg weights
-* Add 1st training of my `gmixer_24_224` MLP /w GLU, 78.1 top-1 w/ 25M params.
-* Add weights from official ResMLP release (https://github.com/facebookresearch/deit)
-* Add `eca_nfnet_l2` weights from my 'lightweight' series. 84.7 top-1 at 384x384.
-* Add distilled BiT 50x1 student and 152x2 Teacher weights from  [Knowledge distillation: A good teacher is patient and consistent](https://arxiv.org/abs/2106.05237)
-* NFNets and ResNetV2-BiT models work w/ Pytorch XLA now
-  * weight standardization uses F.batch_norm instead of std_mean (std_mean wasn't lowered)
-  * eps values adjusted, will be slight differences but should be quite close
-* Improve test coverage and classifier interface of non-conv (vision transformer and mlp) models
-* Cleanup a few classifier / flatten details for models w/ conv classifiers or early global pool
-* Please report any regressions, this PR touched quite a few models.
+	
+二、Docker容器训练：
+    
+1.导入镜像二进制包docker import ubuntuarmpytorch.tar REPOSITORY:TAG, 比如:
 
-### June 8, 2021
-* Add first ResMLP weights, trained in PyTorch XLA on TPU-VM w/ my XLA branch. 24 block variant, 79.2 top-1.
-* Add ResNet51-Q model w/ pretrained weights at 82.36 top-1.
-  * NFNet inspired block layout with quad layer stem and no maxpool
-  * Same param count (35.7M) and throughput as ResNetRS-50 but +1.5 top-1 @ 224x224 and +2.5 top-1 at 288x288
+        docker import ubuntuarmpytorch.tar pytorch:b020
 
-### May 25, 2021
-* Add LeViT, Visformer, ConViT (PR by Aman Arora), Twins (PR by paper authors) transformer models
-* Add ResMLP and gMLP MLP vision models to the existing MLP Mixer impl
-* Fix a number of torchscript issues with various vision transformer models
-* Cleanup input_size/img_size override handling and improve testing / test coverage for all vision transformer and MLP models
-* More flexible pos embedding resize (non-square) for ViT and TnT. Thanks [Alexander Soare](https://github.com/alexander-soare)
-* Add `efficientnetv2_rw_m` model and weights (started training before official code). 84.8 top-1, 53M params.
+2.执行docker_start.sh后带三个参数：步骤1生成的REPOSITORY:TAG；数据集路径；模型执行路径；比如：
 
-### May 14, 2021
-* Add EfficientNet-V2 official model defs w/ ported weights from official [Tensorflow/Keras](https://github.com/google/automl/tree/master/efficientnetv2) impl.
-  * 1k trained variants: `tf_efficientnetv2_s/m/l`
-  * 21k trained variants: `tf_efficientnetv2_s/m/l_in21k`
-  * 21k pretrained -> 1k fine-tuned: `tf_efficientnetv2_s/m/l_in21ft1k`
-  * v2 models w/ v1 scaling: `tf_efficientnetv2_b0` through `b3`
-  * Rename my prev V2 guess `efficientnet_v2s` -> `efficientnetv2_rw_s`
-  * Some blank `efficientnetv2_*` models in-place for future native PyTorch training
+        ./docker_start.sh pytorch:b020 /train/imagenet /home/Gluon_ResNet50_v1c_for_PyTorch
 
-### May 5, 2021
-* Add MLP-Mixer models and port pretrained weights from [Google JAX impl](https://github.com/google-research/vision_transformer/tree/linen)
-* Add CaiT models and pretrained weights from [FB](https://github.com/facebookresearch/deit)
-* Add ResNet-RS models and weights from [TF](https://github.com/tensorflow/tpu/tree/master/models/official/resnet/resnet_rs). Thanks [Aman Arora](https://github.com/amaarora)
-* Add CoaT models and weights. Thanks [Mohammed Rizin](https://github.com/morizin)
-* Add new ImageNet-21k weights & finetuned weights for TResNet, MobileNet-V3, ViT models. Thanks [mrT](https://github.com/mrT23)
-* Add GhostNet models and weights. Thanks [Kai Han](https://github.com/iamhankai)
-* Update ByoaNet attention modules
-   * Improve SA module inits
-   * Hack together experimental stand-alone Swin based attn module and `swinnet`
-   * Consistent '26t' model defs for experiments.
-* Add improved Efficientnet-V2S (prelim model def) weights. 83.8 top-1.
-* WandB logging support
+3.执行步骤一训练流程（环境安装除外）
+	
+三、测试结果
+    
+训练日志路径：在训练脚本的同目录下result文件夹里，如：
 
-### April 13, 2021
-* Add Swin Transformer models and weights from https://github.com/microsoft/Swin-Transformer
-
-### April 12, 2021
-* Add ECA-NFNet-L1 (slimmed down F1 w/ SiLU, 41M params) trained with this code. 84% top-1 @ 320x320. Trained at 256x256.
-* Add EfficientNet-V2S model (unverified model definition) weights. 83.3 top-1 @ 288x288. Only trained single res 224. Working on progressive training.
-* Add ByoaNet model definition (Bring-your-own-attention) w/ SelfAttention block and corresponding SA/SA-like modules and model defs
-  * Lambda Networks - https://arxiv.org/abs/2102.08602
-  * Bottleneck Transformers - https://arxiv.org/abs/2101.11605
-  * Halo Nets - https://arxiv.org/abs/2103.12731
-* Adabelief optimizer contributed by Juntang Zhuang
-
-### April 1, 2021
-* Add snazzy `benchmark.py` script for bulk `timm` model benchmarking of train and/or inference
-* Add Pooling-based Vision Transformer (PiT) models (from https://github.com/naver-ai/pit)
-  * Merged distilled variant into main for torchscript compatibility
-  * Some `timm` cleanup/style tweaks and weights have hub download support
-* Cleanup Vision Transformer (ViT) models
-  * Merge distilled (DeiT) model into main so that torchscript can work
-  * Support updated weight init (defaults to old still) that closer matches original JAX impl (possibly better training from scratch)
-  * Separate hybrid model defs into different file and add several new model defs to fiddle with, support patch_size != 1 for hybrids
-  * Fix fine-tuning num_class changes (PiT and ViT) and pos_embed resizing (Vit) with distilled variants
-  * nn.Sequential for block stack (does not break downstream compat)
-* TnT (Transformer-in-Transformer) models contributed by author (from https://gitee.com/mindspore/mindspore/tree/master/model_zoo/research/cv/TNT)
-* Add RegNetY-160 weights from DeiT teacher model
-* Add new NFNet-L0 w/ SE attn (rename `nfnet_l0b`->`nfnet_l0`) weights 82.75 top-1 @ 288x288
-* Some fixes/improvements for TFDS dataset wrapper
-
-### March 17, 2021
-* Add new ECA-NFNet-L0 (rename `nfnet_l0c`->`eca_nfnet_l0`) weights trained by myself.
-  * 82.6 top-1 @ 288x288, 82.8 @ 320x320, trained at 224x224
-  * Uses SiLU activation, approx 2x faster than `dm_nfnet_f0` and 50% faster than `nfnet_f0s` w/ 1/3 param count
-* Integrate [Hugging Face model hub](https://huggingface.co/models) into timm create_model and default_cfg handling for pretrained weight and config sharing (more on this soon!)
-* Merge HardCoRe NAS models contributed by https://github.com/yoniaflalo
-* Merge PyTorch trained EfficientNet-EL and pruned ES/EL variants contributed by [DeGirum](https://github.com/DeGirum)
-
-
-### March 7, 2021
-* First 0.4.x PyPi release w/ NFNets (& related), ByoB (GPU-Efficient, RepVGG, etc).
-* Change feature extraction for pre-activation nets (NFNets, ResNetV2) to return features before activation.
-* Tested with PyTorch 1.8 release. Updated CI to use 1.8.
-* Benchmarked several arch on RTX 3090, Titan RTX, and V100 across 1.7.1, 1.8, NGC 20.12, and 21.02. Some interesting performance variations to take note of https://gist.github.com/rwightman/bb59f9e245162cee0e38bd66bd8cd77f
-
-### Feb 18, 2021
-* Add pretrained weights and model variants for NFNet-F* models from [DeepMind Haiku impl](https://github.com/deepmind/deepmind-research/tree/master/nfnets).
-  * Models are prefixed with `dm_`. They require SAME padding conv, skipinit enabled, and activation gains applied in act fn.
-  * These models are big, expect to run out of GPU memory. With the GELU activiation + other options, they are roughly 1/2 the inference speed of my SiLU PyTorch optimized `s` variants.
-  * Original model results are based on pre-processing that is not the same as all other models so you'll see different results in the results csv (once updated).
-  * Matching the original pre-processing as closely as possible I get these results:
-    * `dm_nfnet_f6` - 86.352
-    * `dm_nfnet_f5` - 86.100
-    * `dm_nfnet_f4` - 85.834
-    * `dm_nfnet_f3` - 85.676
-    * `dm_nfnet_f2` - 85.178
-    * `dm_nfnet_f1` - 84.696
-    * `dm_nfnet_f0` - 83.464
-
-### Feb 16, 2021
-* Add Adaptive Gradient Clipping (AGC) as per https://arxiv.org/abs/2102.06171. Integrated w/ PyTorch gradient clipping via mode arg that defaults to prev 'norm' mode. For backward arg compat, clip-grad arg must be specified to enable when using train.py.
-  * AGC w/ default clipping factor `--clip-grad .01 --clip-mode agc`
-  * PyTorch global norm of 1.0 (old behaviour, always norm), `--clip-grad 1.0`
-  * PyTorch value clipping of 10, `--clip-grad 10. --clip-mode value`
-  * AGC performance is definitely sensitive to the clipping factor. More experimentation needed to determine good values for smaller batch sizes and optimizers besides those in paper. So far I've found .001-.005 is necessary for stable RMSProp training w/ NFNet/NF-ResNet.
-
-### Feb 12, 2021
-* Update Normalization-Free nets to include new NFNet-F (https://arxiv.org/abs/2102.06171) model defs
-
-### Feb 10, 2021
-* First Normalization-Free model training experiments done,
-  * nf_resnet50 - 80.68 top-1 @ 288x288, 80.31 @ 256x256
-  * nf_regnet_b1 - 79.30 @ 288x288, 78.75 @ 256x256
-* More model archs, incl a flexible ByobNet backbone ('Bring-your-own-blocks')
-  * GPU-Efficient-Networks (https://github.com/idstcv/GPU-Efficient-Networks), impl in `byobnet.py`
-  * RepVGG (https://github.com/DingXiaoH/RepVGG), impl in `byobnet.py`
-  * classic VGG (from torchvision, impl in `vgg.py`)
-* Refinements to normalizer layer arg handling and normalizer+act layer handling in some models
-* Default AMP mode changed to native PyTorch AMP instead of APEX. Issues not being fixed with APEX. Native works with `--channels-last` and `--torchscript` model training, APEX does not.
-* Fix a few bugs introduced since last pypi release
-
-### Feb 8, 2021
-* Add several ResNet weights with ECA attention. 26t & 50t trained @ 256, test @ 320. 269d train @ 256, fine-tune @320, test @ 352.
-  * `ecaresnet26t` - 79.88 top-1 @ 320x320, 79.08 @ 256x256
-  * `ecaresnet50t` - 82.35 top-1 @ 320x320, 81.52 @ 256x256
-  * `ecaresnet269d` - 84.93 top-1 @ 352x352, 84.87 @ 320x320
-* Remove separate tiered (`t`) vs tiered_narrow (`tn`) ResNet model defs, all `tn` changed to `t` and `t` models removed (`seresnext26t_32x4d` only model w/ weights that was removed).
-* Support model default_cfgs with separate train vs test resolution `test_input_size` and remove extra `_320` suffix ResNet model defs that were just for test.
-
-### Jan 30, 2021
-* Add initial "Normalization Free" NF-RegNet-B* and NF-ResNet model definitions based on [paper](https://arxiv.org/abs/2101.08692)
-
-### Jan 25, 2021
-* Add ResNetV2 Big Transfer (BiT) models w/ ImageNet-1k and 21k weights from https://github.com/google-research/big_transfer
-* Add official R50+ViT-B/16 hybrid models + weights from https://github.com/google-research/vision_transformer
-* ImageNet-21k ViT weights are added w/ model defs and representation layer (pre logits) support
-  * NOTE: ImageNet-21k classifier heads were zero'd in original weights, they are only useful for transfer learning
-* Add model defs and weights for DeiT Vision Transformer models from https://github.com/facebookresearch/deit
-* Refactor dataset classes into ImageDataset/IterableImageDataset + dataset specific parser classes
-* Add Tensorflow-Datasets (TFDS) wrapper to allow use of TFDS image classification sets with train script
-  * Ex: `train.py /data/tfds --dataset tfds/oxford_iiit_pet --val-split test --model resnet50 -b 256 --amp --num-classes 37 --opt adamw --lr 3e-4 --weight-decay .001 --pretrained -j 2`
-* Add improved .tar dataset parser that reads images from .tar, folder of .tar files, or .tar within .tar
-  * Run validation on full ImageNet-21k directly from tar w/ BiT model: `validate.py /data/fall11_whole.tar --model resnetv2_50x1_bitm_in21k --amp`
-* Models in this update should be stable w/ possible exception of ViT/BiT, possibility of some regressions with train/val scripts and dataset handling
-
-### Jan 3, 2021
-* Add SE-ResNet-152D weights
-  * 256x256 val, 0.94 crop top-1 - 83.75
-  * 320x320 val, 1.0 crop - 84.36
-* Update [results files](results/)
-
-
-## Introduction
-
-Py**T**orch **Im**age **M**odels (`timm`) is a collection of image models, layers, utilities, optimizers, schedulers, data-loaders / augmentations, and reference training / validation scripts that aim to pull together a wide variety of SOTA models with ability to reproduce ImageNet training results.
-
-The work of many others is present here. I've tried to make sure all source material is acknowledged via links to github, arxiv papers, etc in the README, documentation, and code docstrings. Please let me know if I missed anything.
-
-## Models
-
-All model architecture families include variants with pretrained weights. There are specific model variants without any weights, it is NOT a bug. Help training new or better weights is always appreciated. Here are some example [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) to get you started.
-
-A full version of the list below with source links can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/models/).
-
-* Big Transfer ResNetV2 (BiT) - https://arxiv.org/abs/1912.11370
-* Bottleneck Transformers - https://arxiv.org/abs/2101.11605
-* CaiT (Class-Attention in Image Transformers) - https://arxiv.org/abs/2103.17239
-* CoaT (Co-Scale Conv-Attentional Image Transformers) - https://arxiv.org/abs/2104.06399
-* ConViT (Soft Convolutional Inductive Biases Vision Transformers)- https://arxiv.org/abs/2103.10697
-* CspNet (Cross-Stage Partial Networks) - https://arxiv.org/abs/1911.11929
-* DeiT (Vision Transformer) - https://arxiv.org/abs/2012.12877
-* DenseNet - https://arxiv.org/abs/1608.06993
-* DLA - https://arxiv.org/abs/1707.06484
-* DPN (Dual-Path Network) - https://arxiv.org/abs/1707.01629
-* EfficientNet (MBConvNet Family)
-    * EfficientNet NoisyStudent (B0-B7, L2) - https://arxiv.org/abs/1911.04252
-    * EfficientNet AdvProp (B0-B8) - https://arxiv.org/abs/1911.09665
-    * EfficientNet (B0-B7) - https://arxiv.org/abs/1905.11946
-    * EfficientNet-EdgeTPU (S, M, L) - https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html
-    * EfficientNet V2 - https://arxiv.org/abs/2104.00298
-    * FBNet-C - https://arxiv.org/abs/1812.03443
-    * MixNet - https://arxiv.org/abs/1907.09595
-    * MNASNet B1, A1 (Squeeze-Excite), and Small - https://arxiv.org/abs/1807.11626
-    * MobileNet-V2 - https://arxiv.org/abs/1801.04381
-    * Single-Path NAS - https://arxiv.org/abs/1904.02877
-* GhostNet - https://arxiv.org/abs/1911.11907
-* gMLP - https://arxiv.org/abs/2105.08050
-* GPU-Efficient Networks - https://arxiv.org/abs/2006.14090
-* Halo Nets - https://arxiv.org/abs/2103.12731
-* HardCoRe-NAS - https://arxiv.org/abs/2102.11646
-* HRNet - https://arxiv.org/abs/1908.07919
-* Inception-V3 - https://arxiv.org/abs/1512.00567
-* Inception-ResNet-V2 and Inception-V4 - https://arxiv.org/abs/1602.07261
-* Lambda Networks - https://arxiv.org/abs/2102.08602
-* LeViT (Vision Transformer in ConvNet's Clothing) - https://arxiv.org/abs/2104.01136
-* MLP-Mixer - https://arxiv.org/abs/2105.01601
-* MobileNet-V3 (MBConvNet w/ Efficient Head) - https://arxiv.org/abs/1905.02244
-* NASNet-A - https://arxiv.org/abs/1707.07012
-* NFNet-F - https://arxiv.org/abs/2102.06171
-* NF-RegNet / NF-ResNet - https://arxiv.org/abs/2101.08692
-* PNasNet - https://arxiv.org/abs/1712.00559
-* Pooling-based Vision Transformer (PiT) - https://arxiv.org/abs/2103.16302
-* RegNet - https://arxiv.org/abs/2003.13678
-* RepVGG - https://arxiv.org/abs/2101.03697
-* ResMLP - https://arxiv.org/abs/2105.03404
-* ResNet/ResNeXt
-    * ResNet (v1b/v1.5) - https://arxiv.org/abs/1512.03385
-    * ResNeXt - https://arxiv.org/abs/1611.05431
-    * 'Bag of Tricks' / Gluon C, D, E, S variations - https://arxiv.org/abs/1812.01187
-    * Weakly-supervised (WSL) Instagram pretrained / ImageNet tuned ResNeXt101 - https://arxiv.org/abs/1805.00932
-    * Semi-supervised (SSL) / Semi-weakly Supervised (SWSL) ResNet/ResNeXts - https://arxiv.org/abs/1905.00546
-    * ECA-Net (ECAResNet) - https://arxiv.org/abs/1910.03151v4
-    * Squeeze-and-Excitation Networks (SEResNet) - https://arxiv.org/abs/1709.01507
-    * ResNet-RS - https://arxiv.org/abs/2103.07579
-* Res2Net - https://arxiv.org/abs/1904.01169
-* ResNeSt - https://arxiv.org/abs/2004.08955
-* ReXNet - https://arxiv.org/abs/2007.00992
-* SelecSLS - https://arxiv.org/abs/1907.00837
-* Selective Kernel Networks - https://arxiv.org/abs/1903.06586
-* Swin Transformer - https://arxiv.org/abs/2103.14030
-* Transformer-iN-Transformer (TNT) - https://arxiv.org/abs/2103.00112
-* TResNet - https://arxiv.org/abs/2003.13630
-* Twins (Spatial Attention in Vision Transformers) - https://arxiv.org/pdf/2104.13840.pdf
-* Vision Transformer - https://arxiv.org/abs/2010.11929
-* VovNet V2 and V1 - https://arxiv.org/abs/1911.06667
-* Xception - https://arxiv.org/abs/1610.02357
-* Xception (Modified Aligned, Gluon) - https://arxiv.org/abs/1802.02611
-* Xception (Modified Aligned, TF) - https://arxiv.org/abs/1802.02611
-
-## Features
-
-Several (less common) features that I often utilize in my projects are included. Many of their additions are the reason why I maintain my own set of models, instead of using others' via PIP:
-
-* All models have a common default configuration interface and API for
-    * accessing/changing the classifier - `get_classifier` and `reset_classifier`
-    * doing a forward pass on just the features - `forward_features` (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
-    * these makes it easy to write consistent network wrappers that work with any of the models
-* All models support multi-scale feature map extraction (feature pyramids) via create_model (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
-    * `create_model(name, features_only=True, out_indices=..., output_stride=...)`
-    * `out_indices` creation arg specifies which feature maps to return, these indices are 0 based and generally correspond to the `C(i + 1)` feature level.
-    * `output_stride` creation arg controls output stride of the network by using dilated convolutions. Most networks are stride 32 by default. Not all networks support this.
-    * feature map channel counts, reduction level (stride) can be queried AFTER model creation via the `.feature_info` member
-* All models have a consistent pretrained weight loader that adapts last linear if necessary, and from 3 to 1 channel input if desired
-* High performance [reference training, validation, and inference scripts](https://rwightman.github.io/pytorch-image-models/scripts/) that work in several process/GPU modes:
-    * NVIDIA DDP w/ a single GPU per process, multiple processes with APEX present (AMP mixed-precision optional)
-    * PyTorch DistributedDataParallel w/ multi-gpu, single process (AMP disabled as it crashes when enabled)
-    * PyTorch w/ single GPU single process (AMP optional)
-* A dynamic global pool implementation that allows selecting from average pooling, max pooling, average + max, or concat([average, max]) at model creation. All global pooling is adaptive average by default and compatible with pretrained weights.
-* A 'Test Time Pool' wrapper that can wrap any of the included models and usually provides improved performance doing inference with input images larger than the training size. Idea adapted from original DPN implementation when I ported (https://github.com/cypw/DPNs)
-* Learning rate schedulers
-  * Ideas adopted from
-     * [AllenNLP schedulers](https://github.com/allenai/allennlp/tree/master/allennlp/training/learning_rate_schedulers)
-     * [FAIRseq lr_scheduler](https://github.com/pytorch/fairseq/tree/master/fairseq/optim/lr_scheduler)
-     * SGDR: Stochastic Gradient Descent with Warm Restarts (https://arxiv.org/abs/1608.03983)
-  * Schedulers include `step`, `cosine` w/ restarts, `tanh` w/ restarts, `plateau`
-* Optimizers:
-    * `rmsprop_tf` adapted from PyTorch RMSProp by myself. Reproduces much improved Tensorflow RMSProp behaviour.
-    * `radam` by [Liyuan Liu](https://github.com/LiyuanLucasLiu/RAdam) (https://arxiv.org/abs/1908.03265)
-    * `novograd` by [Masashi Kimura](https://github.com/convergence-lab/novograd) (https://arxiv.org/abs/1905.11286)
-    * `lookahead` adapted from impl by [Liam](https://github.com/alphadl/lookahead.pytorch) (https://arxiv.org/abs/1907.08610)
-    * `fused<name>` optimizers by name with [NVIDIA Apex](https://github.com/NVIDIA/apex/tree/master/apex/optimizers) installed
-    * `adamp` and `sgdp` by [Naver ClovAI](https://github.com/clovaai) (https://arxiv.org/abs/2006.08217)
-    * `adafactor` adapted from [FAIRSeq impl](https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py) (https://arxiv.org/abs/1804.04235)
-    * `adahessian` by [David Samuel](https://github.com/davda54/ada-hessian) (https://arxiv.org/abs/2006.00719)
-* Random Erasing from [Zhun Zhong](https://github.com/zhunzhong07/Random-Erasing/blob/master/transforms.py)  (https://arxiv.org/abs/1708.04896)
-* Mixup (https://arxiv.org/abs/1710.09412)
-* CutMix (https://arxiv.org/abs/1905.04899)
-* AutoAugment (https://arxiv.org/abs/1805.09501) and RandAugment (https://arxiv.org/abs/1909.13719) ImageNet configurations modeled after impl for EfficientNet training (https://github.com/tensorflow/tpu/blob/master/models/official/efficientnet/autoaugment.py)
-* AugMix w/ JSD loss (https://arxiv.org/abs/1912.02781), JSD w/ clean + augmented mixing support works with AutoAugment and RandAugment as well
-* SplitBachNorm - allows splitting batch norm layers between clean and augmented (auxiliary batch norm) data
-* DropPath aka "Stochastic Depth" (https://arxiv.org/abs/1603.09382) 
-* DropBlock (https://arxiv.org/abs/1810.12890)
-* Blur Pooling (https://arxiv.org/abs/1904.11486)
-* Space-to-Depth by [mrT23](https://github.com/mrT23/TResNet/blob/master/src/models/tresnet/layers/space_to_depth.py) (https://arxiv.org/abs/1801.04590) -- original paper?
-* Adaptive Gradient Clipping (https://arxiv.org/abs/2102.06171, https://github.com/deepmind/deepmind-research/tree/master/nfnets)
-* An extensive selection of channel and/or spatial attention modules:
-    * Bottleneck Transformer - https://arxiv.org/abs/2101.11605
-    * CBAM - https://arxiv.org/abs/1807.06521
-    * Effective Squeeze-Excitation (ESE) - https://arxiv.org/abs/1911.06667
-    * Efficient Channel Attention (ECA) - https://arxiv.org/abs/1910.03151
-    * Gather-Excite (GE) - https://arxiv.org/abs/1810.12348
-    * Global Context (GC) - https://arxiv.org/abs/1904.11492
-    * Halo - https://arxiv.org/abs/2103.12731
-    * Involution - https://arxiv.org/abs/2103.06255
-    * Lambda Layer - https://arxiv.org/abs/2102.08602
-    * Non-Local (NL) -  https://arxiv.org/abs/1711.07971
-    * Squeeze-and-Excitation (SE) - https://arxiv.org/abs/1709.01507
-    * Selective Kernel (SK) - (https://arxiv.org/abs/1903.06586
-    * Split (SPLAT) - https://arxiv.org/abs/2004.08955
-    * Shifted Window (SWIN) - https://arxiv.org/abs/2103.14030
-
-## Results
-
-Model validation results can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/results/) and in the [results tables](results/README.md)
-
-## Getting Started (Documentation)
-
-My current [documentation](https://rwightman.github.io/pytorch-image-models/) for `timm` covers the basics.
-
-[timmdocs](https://fastai.github.io/timmdocs/) is quickly becoming a much more comprehensive set of documentation for `timm`. A big thanks to [Aman Arora](https://github.com/amaarora) for his efforts creating timmdocs.
-
-[paperswithcode](https://paperswithcode.com/lib/timm) is a good resource for browsing the models within `timm`.
-
-## Train, Validation, Inference Scripts
-
-The root folder of the repository contains reference train, validation, and inference scripts that work with the included models and other features of this repository. They are adaptable for other datasets and use cases with a little hacking. See [documentation](https://rwightman.github.io/pytorch-image-models/scripts/) for some basics and [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) for some train examples that produce SOTA ImageNet results.
-
-## Awesome PyTorch Resources
-
-One of the greatest assets of PyTorch is the community and their contributions. A few of my favourite resources that pair well with the models and components here are listed below.
-
-### Object Detection, Instance and Semantic Segmentation
-* Detectron2 - https://github.com/facebookresearch/detectron2
-* Segmentation Models (Semantic) - https://github.com/qubvel/segmentation_models.pytorch
-* EfficientDet (Obj Det, Semantic soon) - https://github.com/rwightman/efficientdet-pytorch
-
-### Computer Vision / Image Augmentation
-* Albumentations - https://github.com/albumentations-team/albumentations
-* Kornia - https://github.com/kornia/kornia
-
-### Knowledge Distillation
-* RepDistiller - https://github.com/HobbitLong/RepDistiller
-* torchdistill - https://github.com/yoshitomo-matsubara/torchdistill
-
-### Metric Learning
-* PyTorch Metric Learning - https://github.com/KevinMusgrave/pytorch-metric-learning
-
-### Training / Frameworks
-* fastai - https://github.com/fastai/fastai
-
-## Licenses
-
-### Code
-The code here is licensed Apache 2.0. I've taken care to make sure any third party code included or adapted has compatible (permissive) licenses such as MIT, BSD, etc. I've made an effort to avoid any GPL / LGPL conflicts. That said, it is your responsibility to ensure you comply with licenses here and conditions of any dependent licenses. Where applicable, I've linked the sources/references for various components in docstrings. If you think I've missed anything please create an issue.
-
-### Pretrained Weights
-So far all of the pretrained weights available here are pretrained on ImageNet with a select few that have some additional pretraining (see extra note below). ImageNet was released for non-commercial research purposes only (http://www.image-net.org/download-faq). It's not clear what the implications of that are for the use of pretrained weights from that dataset. Any models I have trained with ImageNet are done for research purposes and one should assume that the original dataset license applies to the weights. It's best to seek legal advice if you intend to use the pretrained weights in a commercial product.
-
-#### Pretrained on more than ImageNet
-Several weights included or references here were pretrained with proprietary datasets that I do not have access to. These include the Facebook WSL, SSL, SWSL ResNe(Xt) and the Google Noisy Student EfficientNet models. The Facebook models have an explicit non-commercial license (CC-BY-NC 4.0, https://github.com/facebookresearch/semi-supervised-ImageNet1K-models, https://github.com/facebookresearch/WSL-Images). The Google models do not appear to have any restriction beyond the Apache 2.0 license (and ImageNet concerns). In either case, you should contact Facebook or Google with any questions.
-
-## Citing
-
-### BibTeX
-
-```
-@misc{rw2019timm,
-  author = {Ross Wightman},
-  title = {PyTorch Image Models},
-  year = {2019},
-  publisher = {GitHub},
-  journal = {GitHub repository},
-  doi = {10.5281/zenodo.4414861},
-  howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
-}
-```
-
-### Latest DOI
-
-[![DOI](https://zenodo.org/badge/168799526.svg)](https://zenodo.org/badge/latestdoi/168799526)
+        /home/Gluon_ResNet50_v1c_for_PyTorch/test/output/0
\ No newline at end of file
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README_raw.md b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README_raw.md
new file mode 100644
index 0000000000..76261cccf2
--- /dev/null
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/README_raw.md
@@ -0,0 +1,399 @@
+# PyTorch Image Models
+- [Sponsors](#sponsors)
+- [What's New](#whats-new)
+- [Introduction](#introduction)
+- [Models](#models)
+- [Features](#features)
+- [Results](#results)
+- [Getting Started (Documentation)](#getting-started-documentation)
+- [Train, Validation, Inference Scripts](#train-validation-inference-scripts)
+- [Awesome PyTorch Resources](#awesome-pytorch-resources)
+- [Licenses](#licenses)
+- [Citing](#citing)
+
+## Sponsors
+
+A big thank you to my [GitHub Sponsors](https://github.com/sponsors/rwightman) for their support!
+
+In addition to the sponsors at the link above, I've received hardware and/or cloud resources from
+* Nvidia (https://www.nvidia.com/en-us/)
+* TFRC (https://www.tensorflow.org/tfrc)
+
+I'm fortunate to be able to dedicate significant time and money of my own supporting this and other open source projects. However, as the projects increase in scope, outside support is needed to continue with the current trajectory of hardware, infrastructure, and electricty costs.
+
+## What's New
+
+### June 20, 2021
+* Release Vision Transformer 'AugReg' weights from [How to train your ViT? Data, Augmentation, and Regularization in Vision Transformers](https://arxiv.org/abs/2106.10270)
+  * .npz weight loading support added, can load any of the 50K+ weights from the [AugReg series](https://console.cloud.google.com/storage/browser/vit_models/augreg)
+  * See [example notebook](https://colab.research.google.com/github/google-research/vision_transformer/blob/master/vit_jax_augreg.ipynb) from official impl for navigating the augreg weights
+  * Replaced all default weights w/ best AugReg variant (if possible). All AugReg 21k classifiers work.
+    * Highlights: `vit_large_patch16_384` (87.1 top-1), `vit_large_r50_s32_384` (86.2 top-1), `vit_base_patch16_384` (86.0 top-1)
+  * `vit_deit_*` renamed to just `deit_*`
+  * Remove my old small model, replace with DeiT compatible small w/ AugReg weights
+* Add 1st training of my `gmixer_24_224` MLP /w GLU, 78.1 top-1 w/ 25M params.
+* Add weights from official ResMLP release (https://github.com/facebookresearch/deit)
+* Add `eca_nfnet_l2` weights from my 'lightweight' series. 84.7 top-1 at 384x384.
+* Add distilled BiT 50x1 student and 152x2 Teacher weights from  [Knowledge distillation: A good teacher is patient and consistent](https://arxiv.org/abs/2106.05237)
+* NFNets and ResNetV2-BiT models work w/ Pytorch XLA now
+  * weight standardization uses F.batch_norm instead of std_mean (std_mean wasn't lowered)
+  * eps values adjusted, will be slight differences but should be quite close
+* Improve test coverage and classifier interface of non-conv (vision transformer and mlp) models
+* Cleanup a few classifier / flatten details for models w/ conv classifiers or early global pool
+* Please report any regressions, this PR touched quite a few models.
+
+### June 8, 2021
+* Add first ResMLP weights, trained in PyTorch XLA on TPU-VM w/ my XLA branch. 24 block variant, 79.2 top-1.
+* Add ResNet51-Q model w/ pretrained weights at 82.36 top-1.
+  * NFNet inspired block layout with quad layer stem and no maxpool
+  * Same param count (35.7M) and throughput as ResNetRS-50 but +1.5 top-1 @ 224x224 and +2.5 top-1 at 288x288
+
+### May 25, 2021
+* Add LeViT, Visformer, ConViT (PR by Aman Arora), Twins (PR by paper authors) transformer models
+* Add ResMLP and gMLP MLP vision models to the existing MLP Mixer impl
+* Fix a number of torchscript issues with various vision transformer models
+* Cleanup input_size/img_size override handling and improve testing / test coverage for all vision transformer and MLP models
+* More flexible pos embedding resize (non-square) for ViT and TnT. Thanks [Alexander Soare](https://github.com/alexander-soare)
+* Add `efficientnetv2_rw_m` model and weights (started training before official code). 84.8 top-1, 53M params.
+
+### May 14, 2021
+* Add EfficientNet-V2 official model defs w/ ported weights from official [Tensorflow/Keras](https://github.com/google/automl/tree/master/efficientnetv2) impl.
+  * 1k trained variants: `tf_efficientnetv2_s/m/l`
+  * 21k trained variants: `tf_efficientnetv2_s/m/l_in21k`
+  * 21k pretrained -> 1k fine-tuned: `tf_efficientnetv2_s/m/l_in21ft1k`
+  * v2 models w/ v1 scaling: `tf_efficientnetv2_b0` through `b3`
+  * Rename my prev V2 guess `efficientnet_v2s` -> `efficientnetv2_rw_s`
+  * Some blank `efficientnetv2_*` models in-place for future native PyTorch training
+
+### May 5, 2021
+* Add MLP-Mixer models and port pretrained weights from [Google JAX impl](https://github.com/google-research/vision_transformer/tree/linen)
+* Add CaiT models and pretrained weights from [FB](https://github.com/facebookresearch/deit)
+* Add ResNet-RS models and weights from [TF](https://github.com/tensorflow/tpu/tree/master/models/official/resnet/resnet_rs). Thanks [Aman Arora](https://github.com/amaarora)
+* Add CoaT models and weights. Thanks [Mohammed Rizin](https://github.com/morizin)
+* Add new ImageNet-21k weights & finetuned weights for TResNet, MobileNet-V3, ViT models. Thanks [mrT](https://github.com/mrT23)
+* Add GhostNet models and weights. Thanks [Kai Han](https://github.com/iamhankai)
+* Update ByoaNet attention modules
+   * Improve SA module inits
+   * Hack together experimental stand-alone Swin based attn module and `swinnet`
+   * Consistent '26t' model defs for experiments.
+* Add improved Efficientnet-V2S (prelim model def) weights. 83.8 top-1.
+* WandB logging support
+
+### April 13, 2021
+* Add Swin Transformer models and weights from https://github.com/microsoft/Swin-Transformer
+
+### April 12, 2021
+* Add ECA-NFNet-L1 (slimmed down F1 w/ SiLU, 41M params) trained with this code. 84% top-1 @ 320x320. Trained at 256x256.
+* Add EfficientNet-V2S model (unverified model definition) weights. 83.3 top-1 @ 288x288. Only trained single res 224. Working on progressive training.
+* Add ByoaNet model definition (Bring-your-own-attention) w/ SelfAttention block and corresponding SA/SA-like modules and model defs
+  * Lambda Networks - https://arxiv.org/abs/2102.08602
+  * Bottleneck Transformers - https://arxiv.org/abs/2101.11605
+  * Halo Nets - https://arxiv.org/abs/2103.12731
+* Adabelief optimizer contributed by Juntang Zhuang
+
+### April 1, 2021
+* Add snazzy `benchmark.py` script for bulk `timm` model benchmarking of train and/or inference
+* Add Pooling-based Vision Transformer (PiT) models (from https://github.com/naver-ai/pit)
+  * Merged distilled variant into main for torchscript compatibility
+  * Some `timm` cleanup/style tweaks and weights have hub download support
+* Cleanup Vision Transformer (ViT) models
+  * Merge distilled (DeiT) model into main so that torchscript can work
+  * Support updated weight init (defaults to old still) that closer matches original JAX impl (possibly better training from scratch)
+  * Separate hybrid model defs into different file and add several new model defs to fiddle with, support patch_size != 1 for hybrids
+  * Fix fine-tuning num_class changes (PiT and ViT) and pos_embed resizing (Vit) with distilled variants
+  * nn.Sequential for block stack (does not break downstream compat)
+* TnT (Transformer-in-Transformer) models contributed by author (from https://gitee.com/mindspore/mindspore/tree/master/model_zoo/research/cv/TNT)
+* Add RegNetY-160 weights from DeiT teacher model
+* Add new NFNet-L0 w/ SE attn (rename `nfnet_l0b`->`nfnet_l0`) weights 82.75 top-1 @ 288x288
+* Some fixes/improvements for TFDS dataset wrapper
+
+### March 17, 2021
+* Add new ECA-NFNet-L0 (rename `nfnet_l0c`->`eca_nfnet_l0`) weights trained by myself.
+  * 82.6 top-1 @ 288x288, 82.8 @ 320x320, trained at 224x224
+  * Uses SiLU activation, approx 2x faster than `dm_nfnet_f0` and 50% faster than `nfnet_f0s` w/ 1/3 param count
+* Integrate [Hugging Face model hub](https://huggingface.co/models) into timm create_model and default_cfg handling for pretrained weight and config sharing (more on this soon!)
+* Merge HardCoRe NAS models contributed by https://github.com/yoniaflalo
+* Merge PyTorch trained EfficientNet-EL and pruned ES/EL variants contributed by [DeGirum](https://github.com/DeGirum)
+
+
+### March 7, 2021
+* First 0.4.x PyPi release w/ NFNets (& related), ByoB (GPU-Efficient, RepVGG, etc).
+* Change feature extraction for pre-activation nets (NFNets, ResNetV2) to return features before activation.
+* Tested with PyTorch 1.8 release. Updated CI to use 1.8.
+* Benchmarked several arch on RTX 3090, Titan RTX, and V100 across 1.7.1, 1.8, NGC 20.12, and 21.02. Some interesting performance variations to take note of https://gist.github.com/rwightman/bb59f9e245162cee0e38bd66bd8cd77f
+
+### Feb 18, 2021
+* Add pretrained weights and model variants for NFNet-F* models from [DeepMind Haiku impl](https://github.com/deepmind/deepmind-research/tree/master/nfnets).
+  * Models are prefixed with `dm_`. They require SAME padding conv, skipinit enabled, and activation gains applied in act fn.
+  * These models are big, expect to run out of GPU memory. With the GELU activiation + other options, they are roughly 1/2 the inference speed of my SiLU PyTorch optimized `s` variants.
+  * Original model results are based on pre-processing that is not the same as all other models so you'll see different results in the results csv (once updated).
+  * Matching the original pre-processing as closely as possible I get these results:
+    * `dm_nfnet_f6` - 86.352
+    * `dm_nfnet_f5` - 86.100
+    * `dm_nfnet_f4` - 85.834
+    * `dm_nfnet_f3` - 85.676
+    * `dm_nfnet_f2` - 85.178
+    * `dm_nfnet_f1` - 84.696
+    * `dm_nfnet_f0` - 83.464
+
+### Feb 16, 2021
+* Add Adaptive Gradient Clipping (AGC) as per https://arxiv.org/abs/2102.06171. Integrated w/ PyTorch gradient clipping via mode arg that defaults to prev 'norm' mode. For backward arg compat, clip-grad arg must be specified to enable when using train.py.
+  * AGC w/ default clipping factor `--clip-grad .01 --clip-mode agc`
+  * PyTorch global norm of 1.0 (old behaviour, always norm), `--clip-grad 1.0`
+  * PyTorch value clipping of 10, `--clip-grad 10. --clip-mode value`
+  * AGC performance is definitely sensitive to the clipping factor. More experimentation needed to determine good values for smaller batch sizes and optimizers besides those in paper. So far I've found .001-.005 is necessary for stable RMSProp training w/ NFNet/NF-ResNet.
+
+### Feb 12, 2021
+* Update Normalization-Free nets to include new NFNet-F (https://arxiv.org/abs/2102.06171) model defs
+
+### Feb 10, 2021
+* First Normalization-Free model training experiments done,
+  * nf_resnet50 - 80.68 top-1 @ 288x288, 80.31 @ 256x256
+  * nf_regnet_b1 - 79.30 @ 288x288, 78.75 @ 256x256
+* More model archs, incl a flexible ByobNet backbone ('Bring-your-own-blocks')
+  * GPU-Efficient-Networks (https://github.com/idstcv/GPU-Efficient-Networks), impl in `byobnet.py`
+  * RepVGG (https://github.com/DingXiaoH/RepVGG), impl in `byobnet.py`
+  * classic VGG (from torchvision, impl in `vgg.py`)
+* Refinements to normalizer layer arg handling and normalizer+act layer handling in some models
+* Default AMP mode changed to native PyTorch AMP instead of APEX. Issues not being fixed with APEX. Native works with `--channels-last` and `--torchscript` model training, APEX does not.
+* Fix a few bugs introduced since last pypi release
+
+### Feb 8, 2021
+* Add several ResNet weights with ECA attention. 26t & 50t trained @ 256, test @ 320. 269d train @ 256, fine-tune @320, test @ 352.
+  * `ecaresnet26t` - 79.88 top-1 @ 320x320, 79.08 @ 256x256
+  * `ecaresnet50t` - 82.35 top-1 @ 320x320, 81.52 @ 256x256
+  * `ecaresnet269d` - 84.93 top-1 @ 352x352, 84.87 @ 320x320
+* Remove separate tiered (`t`) vs tiered_narrow (`tn`) ResNet model defs, all `tn` changed to `t` and `t` models removed (`seresnext26t_32x4d` only model w/ weights that was removed).
+* Support model default_cfgs with separate train vs test resolution `test_input_size` and remove extra `_320` suffix ResNet model defs that were just for test.
+
+### Jan 30, 2021
+* Add initial "Normalization Free" NF-RegNet-B* and NF-ResNet model definitions based on [paper](https://arxiv.org/abs/2101.08692)
+
+### Jan 25, 2021
+* Add ResNetV2 Big Transfer (BiT) models w/ ImageNet-1k and 21k weights from https://github.com/google-research/big_transfer
+* Add official R50+ViT-B/16 hybrid models + weights from https://github.com/google-research/vision_transformer
+* ImageNet-21k ViT weights are added w/ model defs and representation layer (pre logits) support
+  * NOTE: ImageNet-21k classifier heads were zero'd in original weights, they are only useful for transfer learning
+* Add model defs and weights for DeiT Vision Transformer models from https://github.com/facebookresearch/deit
+* Refactor dataset classes into ImageDataset/IterableImageDataset + dataset specific parser classes
+* Add Tensorflow-Datasets (TFDS) wrapper to allow use of TFDS image classification sets with train script
+  * Ex: `train.py /data/tfds --dataset tfds/oxford_iiit_pet --val-split test --model resnet50 -b 256 --amp --num-classes 37 --opt adamw --lr 3e-4 --weight-decay .001 --pretrained -j 2`
+* Add improved .tar dataset parser that reads images from .tar, folder of .tar files, or .tar within .tar
+  * Run validation on full ImageNet-21k directly from tar w/ BiT model: `validate.py /data/fall11_whole.tar --model resnetv2_50x1_bitm_in21k --amp`
+* Models in this update should be stable w/ possible exception of ViT/BiT, possibility of some regressions with train/val scripts and dataset handling
+
+### Jan 3, 2021
+* Add SE-ResNet-152D weights
+  * 256x256 val, 0.94 crop top-1 - 83.75
+  * 320x320 val, 1.0 crop - 84.36
+* Update [results files](results/)
+
+
+## Introduction
+
+Py**T**orch **Im**age **M**odels (`timm`) is a collection of image models, layers, utilities, optimizers, schedulers, data-loaders / augmentations, and reference training / validation scripts that aim to pull together a wide variety of SOTA models with ability to reproduce ImageNet training results.
+
+The work of many others is present here. I've tried to make sure all source material is acknowledged via links to github, arxiv papers, etc in the README, documentation, and code docstrings. Please let me know if I missed anything.
+
+## Models
+
+All model architecture families include variants with pretrained weights. There are specific model variants without any weights, it is NOT a bug. Help training new or better weights is always appreciated. Here are some example [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) to get you started.
+
+A full version of the list below with source links can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/models/).
+
+* Big Transfer ResNetV2 (BiT) - https://arxiv.org/abs/1912.11370
+* Bottleneck Transformers - https://arxiv.org/abs/2101.11605
+* CaiT (Class-Attention in Image Transformers) - https://arxiv.org/abs/2103.17239
+* CoaT (Co-Scale Conv-Attentional Image Transformers) - https://arxiv.org/abs/2104.06399
+* ConViT (Soft Convolutional Inductive Biases Vision Transformers)- https://arxiv.org/abs/2103.10697
+* CspNet (Cross-Stage Partial Networks) - https://arxiv.org/abs/1911.11929
+* DeiT (Vision Transformer) - https://arxiv.org/abs/2012.12877
+* DenseNet - https://arxiv.org/abs/1608.06993
+* DLA - https://arxiv.org/abs/1707.06484
+* DPN (Dual-Path Network) - https://arxiv.org/abs/1707.01629
+* EfficientNet (MBConvNet Family)
+    * EfficientNet NoisyStudent (B0-B7, L2) - https://arxiv.org/abs/1911.04252
+    * EfficientNet AdvProp (B0-B8) - https://arxiv.org/abs/1911.09665
+    * EfficientNet (B0-B7) - https://arxiv.org/abs/1905.11946
+    * EfficientNet-EdgeTPU (S, M, L) - https://ai.googleblog.com/2019/08/efficientnet-edgetpu-creating.html
+    * EfficientNet V2 - https://arxiv.org/abs/2104.00298
+    * FBNet-C - https://arxiv.org/abs/1812.03443
+    * MixNet - https://arxiv.org/abs/1907.09595
+    * MNASNet B1, A1 (Squeeze-Excite), and Small - https://arxiv.org/abs/1807.11626
+    * MobileNet-V2 - https://arxiv.org/abs/1801.04381
+    * Single-Path NAS - https://arxiv.org/abs/1904.02877
+* GhostNet - https://arxiv.org/abs/1911.11907
+* gMLP - https://arxiv.org/abs/2105.08050
+* GPU-Efficient Networks - https://arxiv.org/abs/2006.14090
+* Halo Nets - https://arxiv.org/abs/2103.12731
+* HardCoRe-NAS - https://arxiv.org/abs/2102.11646
+* HRNet - https://arxiv.org/abs/1908.07919
+* Inception-V3 - https://arxiv.org/abs/1512.00567
+* Inception-ResNet-V2 and Inception-V4 - https://arxiv.org/abs/1602.07261
+* Lambda Networks - https://arxiv.org/abs/2102.08602
+* LeViT (Vision Transformer in ConvNet's Clothing) - https://arxiv.org/abs/2104.01136
+* MLP-Mixer - https://arxiv.org/abs/2105.01601
+* MobileNet-V3 (MBConvNet w/ Efficient Head) - https://arxiv.org/abs/1905.02244
+* NASNet-A - https://arxiv.org/abs/1707.07012
+* NFNet-F - https://arxiv.org/abs/2102.06171
+* NF-RegNet / NF-ResNet - https://arxiv.org/abs/2101.08692
+* PNasNet - https://arxiv.org/abs/1712.00559
+* Pooling-based Vision Transformer (PiT) - https://arxiv.org/abs/2103.16302
+* RegNet - https://arxiv.org/abs/2003.13678
+* RepVGG - https://arxiv.org/abs/2101.03697
+* ResMLP - https://arxiv.org/abs/2105.03404
+* ResNet/ResNeXt
+    * ResNet (v1b/v1.5) - https://arxiv.org/abs/1512.03385
+    * ResNeXt - https://arxiv.org/abs/1611.05431
+    * 'Bag of Tricks' / Gluon C, D, E, S variations - https://arxiv.org/abs/1812.01187
+    * Weakly-supervised (WSL) Instagram pretrained / ImageNet tuned ResNeXt101 - https://arxiv.org/abs/1805.00932
+    * Semi-supervised (SSL) / Semi-weakly Supervised (SWSL) ResNet/ResNeXts - https://arxiv.org/abs/1905.00546
+    * ECA-Net (ECAResNet) - https://arxiv.org/abs/1910.03151v4
+    * Squeeze-and-Excitation Networks (SEResNet) - https://arxiv.org/abs/1709.01507
+    * ResNet-RS - https://arxiv.org/abs/2103.07579
+* Res2Net - https://arxiv.org/abs/1904.01169
+* ResNeSt - https://arxiv.org/abs/2004.08955
+* ReXNet - https://arxiv.org/abs/2007.00992
+* SelecSLS - https://arxiv.org/abs/1907.00837
+* Selective Kernel Networks - https://arxiv.org/abs/1903.06586
+* Swin Transformer - https://arxiv.org/abs/2103.14030
+* Transformer-iN-Transformer (TNT) - https://arxiv.org/abs/2103.00112
+* TResNet - https://arxiv.org/abs/2003.13630
+* Twins (Spatial Attention in Vision Transformers) - https://arxiv.org/pdf/2104.13840.pdf
+* Vision Transformer - https://arxiv.org/abs/2010.11929
+* VovNet V2 and V1 - https://arxiv.org/abs/1911.06667
+* Xception - https://arxiv.org/abs/1610.02357
+* Xception (Modified Aligned, Gluon) - https://arxiv.org/abs/1802.02611
+* Xception (Modified Aligned, TF) - https://arxiv.org/abs/1802.02611
+
+## Features
+
+Several (less common) features that I often utilize in my projects are included. Many of their additions are the reason why I maintain my own set of models, instead of using others' via PIP:
+
+* All models have a common default configuration interface and API for
+    * accessing/changing the classifier - `get_classifier` and `reset_classifier`
+    * doing a forward pass on just the features - `forward_features` (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
+    * these makes it easy to write consistent network wrappers that work with any of the models
+* All models support multi-scale feature map extraction (feature pyramids) via create_model (see [documentation](https://rwightman.github.io/pytorch-image-models/feature_extraction/))
+    * `create_model(name, features_only=True, out_indices=..., output_stride=...)`
+    * `out_indices` creation arg specifies which feature maps to return, these indices are 0 based and generally correspond to the `C(i + 1)` feature level.
+    * `output_stride` creation arg controls output stride of the network by using dilated convolutions. Most networks are stride 32 by default. Not all networks support this.
+    * feature map channel counts, reduction level (stride) can be queried AFTER model creation via the `.feature_info` member
+* All models have a consistent pretrained weight loader that adapts last linear if necessary, and from 3 to 1 channel input if desired
+* High performance [reference training, validation, and inference scripts](https://rwightman.github.io/pytorch-image-models/scripts/) that work in several process/GPU modes:
+    * NVIDIA DDP w/ a single GPU per process, multiple processes with APEX present (AMP mixed-precision optional)
+    * PyTorch DistributedDataParallel w/ multi-gpu, single process (AMP disabled as it crashes when enabled)
+    * PyTorch w/ single GPU single process (AMP optional)
+* A dynamic global pool implementation that allows selecting from average pooling, max pooling, average + max, or concat([average, max]) at model creation. All global pooling is adaptive average by default and compatible with pretrained weights.
+* A 'Test Time Pool' wrapper that can wrap any of the included models and usually provides improved performance doing inference with input images larger than the training size. Idea adapted from original DPN implementation when I ported (https://github.com/cypw/DPNs)
+* Learning rate schedulers
+  * Ideas adopted from
+     * [AllenNLP schedulers](https://github.com/allenai/allennlp/tree/master/allennlp/training/learning_rate_schedulers)
+     * [FAIRseq lr_scheduler](https://github.com/pytorch/fairseq/tree/master/fairseq/optim/lr_scheduler)
+     * SGDR: Stochastic Gradient Descent with Warm Restarts (https://arxiv.org/abs/1608.03983)
+  * Schedulers include `step`, `cosine` w/ restarts, `tanh` w/ restarts, `plateau`
+* Optimizers:
+    * `rmsprop_tf` adapted from PyTorch RMSProp by myself. Reproduces much improved Tensorflow RMSProp behaviour.
+    * `radam` by [Liyuan Liu](https://github.com/LiyuanLucasLiu/RAdam) (https://arxiv.org/abs/1908.03265)
+    * `novograd` by [Masashi Kimura](https://github.com/convergence-lab/novograd) (https://arxiv.org/abs/1905.11286)
+    * `lookahead` adapted from impl by [Liam](https://github.com/alphadl/lookahead.pytorch) (https://arxiv.org/abs/1907.08610)
+    * `fused<name>` optimizers by name with [NVIDIA Apex](https://github.com/NVIDIA/apex/tree/master/apex/optimizers) installed
+    * `adamp` and `sgdp` by [Naver ClovAI](https://github.com/clovaai) (https://arxiv.org/abs/2006.08217)
+    * `adafactor` adapted from [FAIRSeq impl](https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py) (https://arxiv.org/abs/1804.04235)
+    * `adahessian` by [David Samuel](https://github.com/davda54/ada-hessian) (https://arxiv.org/abs/2006.00719)
+* Random Erasing from [Zhun Zhong](https://github.com/zhunzhong07/Random-Erasing/blob/master/transforms.py)  (https://arxiv.org/abs/1708.04896)
+* Mixup (https://arxiv.org/abs/1710.09412)
+* CutMix (https://arxiv.org/abs/1905.04899)
+* AutoAugment (https://arxiv.org/abs/1805.09501) and RandAugment (https://arxiv.org/abs/1909.13719) ImageNet configurations modeled after impl for EfficientNet training (https://github.com/tensorflow/tpu/blob/master/models/official/efficientnet/autoaugment.py)
+* AugMix w/ JSD loss (https://arxiv.org/abs/1912.02781), JSD w/ clean + augmented mixing support works with AutoAugment and RandAugment as well
+* SplitBachNorm - allows splitting batch norm layers between clean and augmented (auxiliary batch norm) data
+* DropPath aka "Stochastic Depth" (https://arxiv.org/abs/1603.09382) 
+* DropBlock (https://arxiv.org/abs/1810.12890)
+* Blur Pooling (https://arxiv.org/abs/1904.11486)
+* Space-to-Depth by [mrT23](https://github.com/mrT23/TResNet/blob/master/src/models/tresnet/layers/space_to_depth.py) (https://arxiv.org/abs/1801.04590) -- original paper?
+* Adaptive Gradient Clipping (https://arxiv.org/abs/2102.06171, https://github.com/deepmind/deepmind-research/tree/master/nfnets)
+* An extensive selection of channel and/or spatial attention modules:
+    * Bottleneck Transformer - https://arxiv.org/abs/2101.11605
+    * CBAM - https://arxiv.org/abs/1807.06521
+    * Effective Squeeze-Excitation (ESE) - https://arxiv.org/abs/1911.06667
+    * Efficient Channel Attention (ECA) - https://arxiv.org/abs/1910.03151
+    * Gather-Excite (GE) - https://arxiv.org/abs/1810.12348
+    * Global Context (GC) - https://arxiv.org/abs/1904.11492
+    * Halo - https://arxiv.org/abs/2103.12731
+    * Involution - https://arxiv.org/abs/2103.06255
+    * Lambda Layer - https://arxiv.org/abs/2102.08602
+    * Non-Local (NL) -  https://arxiv.org/abs/1711.07971
+    * Squeeze-and-Excitation (SE) - https://arxiv.org/abs/1709.01507
+    * Selective Kernel (SK) - (https://arxiv.org/abs/1903.06586
+    * Split (SPLAT) - https://arxiv.org/abs/2004.08955
+    * Shifted Window (SWIN) - https://arxiv.org/abs/2103.14030
+
+## Results
+
+Model validation results can be found in the [documentation](https://rwightman.github.io/pytorch-image-models/results/) and in the [results tables](results/README.md)
+
+## Getting Started (Documentation)
+
+My current [documentation](https://rwightman.github.io/pytorch-image-models/) for `timm` covers the basics.
+
+[timmdocs](https://fastai.github.io/timmdocs/) is quickly becoming a much more comprehensive set of documentation for `timm`. A big thanks to [Aman Arora](https://github.com/amaarora) for his efforts creating timmdocs.
+
+[paperswithcode](https://paperswithcode.com/lib/timm) is a good resource for browsing the models within `timm`.
+
+## Train, Validation, Inference Scripts
+
+The root folder of the repository contains reference train, validation, and inference scripts that work with the included models and other features of this repository. They are adaptable for other datasets and use cases with a little hacking. See [documentation](https://rwightman.github.io/pytorch-image-models/scripts/) for some basics and [training hparams](https://rwightman.github.io/pytorch-image-models/training_hparam_examples) for some train examples that produce SOTA ImageNet results.
+
+## Awesome PyTorch Resources
+
+One of the greatest assets of PyTorch is the community and their contributions. A few of my favourite resources that pair well with the models and components here are listed below.
+
+### Object Detection, Instance and Semantic Segmentation
+* Detectron2 - https://github.com/facebookresearch/detectron2
+* Segmentation Models (Semantic) - https://github.com/qubvel/segmentation_models.pytorch
+* EfficientDet (Obj Det, Semantic soon) - https://github.com/rwightman/efficientdet-pytorch
+
+### Computer Vision / Image Augmentation
+* Albumentations - https://github.com/albumentations-team/albumentations
+* Kornia - https://github.com/kornia/kornia
+
+### Knowledge Distillation
+* RepDistiller - https://github.com/HobbitLong/RepDistiller
+* torchdistill - https://github.com/yoshitomo-matsubara/torchdistill
+
+### Metric Learning
+* PyTorch Metric Learning - https://github.com/KevinMusgrave/pytorch-metric-learning
+
+### Training / Frameworks
+* fastai - https://github.com/fastai/fastai
+
+## Licenses
+
+### Code
+The code here is licensed Apache 2.0. I've taken care to make sure any third party code included or adapted has compatible (permissive) licenses such as MIT, BSD, etc. I've made an effort to avoid any GPL / LGPL conflicts. That said, it is your responsibility to ensure you comply with licenses here and conditions of any dependent licenses. Where applicable, I've linked the sources/references for various components in docstrings. If you think I've missed anything please create an issue.
+
+### Pretrained Weights
+So far all of the pretrained weights available here are pretrained on ImageNet with a select few that have some additional pretraining (see extra note below). ImageNet was released for non-commercial research purposes only (http://www.image-net.org/download-faq). It's not clear what the implications of that are for the use of pretrained weights from that dataset. Any models I have trained with ImageNet are done for research purposes and one should assume that the original dataset license applies to the weights. It's best to seek legal advice if you intend to use the pretrained weights in a commercial product.
+
+#### Pretrained on more than ImageNet
+Several weights included or references here were pretrained with proprietary datasets that I do not have access to. These include the Facebook WSL, SSL, SWSL ResNe(Xt) and the Google Noisy Student EfficientNet models. The Facebook models have an explicit non-commercial license (CC-BY-NC 4.0, https://github.com/facebookresearch/semi-supervised-ImageNet1K-models, https://github.com/facebookresearch/WSL-Images). The Google models do not appear to have any restriction beyond the Apache 2.0 license (and ImageNet concerns). In either case, you should contact Facebook or Google with any questions.
+
+## Citing
+
+### BibTeX
+
+```
+@misc{rw2019timm,
+  author = {Ross Wightman},
+  title = {PyTorch Image Models},
+  year = {2019},
+  publisher = {GitHub},
+  journal = {GitHub repository},
+  doi = {10.5281/zenodo.4414861},
+  howpublished = {\url{https://github.com/rwightman/pytorch-image-models}}
+}
+```
+
+### Latest DOI
+
+[![DOI](https://zenodo.org/badge/168799526.svg)](https://zenodo.org/badge/latestdoi/168799526)
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_1p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_1p.sh
index 9bbf9a4a54..723dc9f8a8 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_1p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_1p.sh
@@ -74,10 +74,16 @@ fi
 #################启动训练脚本#################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
 
-python3 train.py \
+# 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
+
+
+nohup python3.7 train.py \
     ${data_path} \
     -b ${batch_size} \
     --model gluon_resnet50_v1d \
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_8p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_8p.sh
index 99aa5a232d..e8c85f7a34 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_8p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_full_8p.sh
@@ -63,12 +63,18 @@ fi
 #################启动训练脚本#################
 # 训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
+
+# 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
+
 
 for i in $(seq 0 7)
 do
-    python3 train.py \
+    nohup python3.7 train.py \
         ${data_path} \
         -b ${batch_size} \
         --model gluon_resnet50_v1d \
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_1p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_1p.sh
index 0d1d73aaac..794f9ca5c6 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_1p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_1p.sh
@@ -74,10 +74,16 @@ fi
 #################启动训练脚本#################
 #训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
 
-python3 train.py \
+# 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
+
+
+nohup python3.7 train.py \
     ${data_path} \
     -b ${batch_size} \
     --model gluon_resnet50_v1d \
diff --git a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_8p.sh b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_8p.sh
index 4592db0418..c7514edcde 100644
--- a/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_8p.sh
+++ b/PyTorch/built-in/cv/classification/Gluon_ResNet50_v1d_for_PyTorch/test/train_performance_8p.sh
@@ -63,12 +63,18 @@ fi
 #################启动训练脚本#################
 # 训练开始时间，不需要修改
 start_time=$(date +%s)
-# source 环境变量
-source ${test_path_dir}/env_npu.sh
+
+# 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
+
 
 for i in $(seq 0 7)
 do
-    python3 train.py \
+    nohup python3.7 train.py \
         ${data_path} \
         -b ${batch_size} \
         --model gluon_resnet50_v1d \
-- 
Gitee