add version 1 for benchmark

benchmark/README.md

+# benchmark使用说明
+
+此目录所有shell脚本是为了测试PaddleClas中不同模型的速度指标，如单卡训练速度指标、多卡训练速度指标等。
+
+## 相关脚本说明
+
+一共有3个脚本：
+
+- `prepare_data.sh`: 下载相应的测试数据，并配置好数据路径
+- `run_benchmark.sh`: 执行单独一个训练测试的脚本，具体调用方式，可查看脚本注释
+- `run_all.sh`: 执行所有训练测试的入口脚本
+
+## 使用说明
+
+******为了跟PaddleClas中其他的模块的执行目录保持一致，此模块的执行目录为`PaddleClas`的根目录。
+
+### 1.准备数据
+
+```shell
+bash benchmark/prepare_data.sh
+```
+
+### 2.执行所有模型的测试
+
+```shell
+bash benchmark/run_all.sh
+```

benchmark/prepare_data.sh

+#!/bin/bash
+
+cd dataset
+rm -rf ILSVRC2012
+wget -nc https://paddle-imagenet-models-name.bj.bcebos.com/data/whole_chain/whole_chain_little_train.tar
+tar xf whole_chain_little_train.tar
+ln -s whole_chain_little_train ILSVRC2012
+cd ILSVRC2012 
+mv train.txt train_list.txt
+mv val.txt val_list.txt
+cd ../../

benchmark/run_all.sh

+# 提供可稳定复现性能的脚本，默认在标准docker环境内py37执行： paddlepaddle/paddle:latest-gpu-cuda10.1-cudnn7  paddle=2.1.2  py=37
+# 执行目录：需说明
+# cd **
+# 1 安装该模型需要的依赖 (如需开启优化策略请注明)
+# pip install ...
+# 2 拷贝该模型需要数据、预训练模型
+# 3 批量运行（如不方便批量，1，2需放到单个模型中）
+ 
+model_mode_list=(MobileNetV1 MobileNetV2 MobileNetV3_large_x1_0 EfficientNetB0 ShuffleNetV2_x1_0 DenseNet121 HRNet_W48_C SwinTransformer_tiny_patch4_window7_224 alt_gvt_base)
+fp_item_list=(fp32)
+bs_list=(32 64 96 128)
+for model_mode in ${model_mode_list[@]}; do
+      for fp_item in ${fp_item_list[@]}; do
+          for bs_item in ${bs_list[@]};do
+	    echo "index is speed, 1gpus, begin, ${model_name}"
+	    run_mode=sp
+	    CUDA_VISIBLE_DEVICES=0 bash benchmark/run_benchmark.sh ${run_mode} ${bs_item} ${fp_item} 10 ${model_mode}     #  (5min)
+	    sleep 10
+            echo "index is speed, 8gpus, run_mode is multi_process, begin, ${model_name}"
+            run_mode=mp
+            CUDA_VISIBLE_DEVICES=0,1,2,3,4,5,6,7 bash benchmark/run_benchmark.sh ${run_mode} ${bs_item} ${fp_item} 10 ${model_mode} 
+            sleep 10
+            done
+      done
+done

benchmark/run_benchmark.sh

+#!/usr/bin/env bash
+set -xe
+# 运行示例：CUDA_VISIBLE_DEVICES=0 bash run_benchmark.sh ${run_mode} ${bs_item} ${fp_item} 500 ${model_mode}
+# 参数说明
+function _set_params(){
+    run_mode=${1:-"sp"}          # 单卡sp|多卡mp
+    batch_size=${2:-"64"}
+    fp_item=${3:-"fp32"}        # fp32|fp16
+    epochs=${4:-"10"}       # 可选，如果需要修改代码提前中断
+    model_name=${5:-"model_name"}
+    run_log_path="${TRAIN_LOG_DIR:-$(pwd)}/benchmark"  # TRAIN_LOG_DIR 后续QA设置该参数
+ 
+#   以下不用修改   
+    device=${CUDA_VISIBLE_DEVICES//,/ }
+    arr=(${device})
+    num_gpu_devices=${#arr[*]}
+    log_file=${run_log_path}/${model_name}_${run_mode}_bs${batch_size}_${fp_item}_${num_gpu_devices}
+}
+function _train(){
+    echo "Train on ${num_gpu_devices} GPUs"
+    echo "current CUDA_VISIBLE_DEVICES=$CUDA_VISIBLE_DEVICES, gpus=$num_gpu_devices, batch_size=$batch_size"
+
+    if [ ${fp_item} = "fp32" ];then
+    	model_config=`find ppcls/configs/ -name ${model_name}.yaml` 
+    else
+	model_config=`find ppcls/configs/ -name ${model_name}_fp16.yaml` 
+    fi
+
+    train_cmd="-c ${model_config} -o DataLoader.Train.sampler.batch_size=${batch_size} -o Global.epochs=${epochs}"   
+    case ${run_mode} in
+    sp) train_cmd="python -u tools/train.py ${train_cmd}" ;;
+    mp)
+        train_cmd="python -m paddle.distributed.launch --log_dir=./mylog --gpus=$CUDA_VISIBLE_DEVICES tools/train.py ${train_cmd}"
+        log_parse_file="mylog/workerlog.0" ;;
+    *) echo "choose run_mode(sp or mp)"; exit 1;
+    esac
+# 以下不用修改
+    timeout 15m ${train_cmd} > ${log_file} 2>&1
+    if [ $? -ne 0 ];then
+        echo -e "${model_name}, FAIL"
+        export job_fail_flag=1
+    else
+        echo -e "${model_name}, SUCCESS"
+        export job_fail_flag=0
+    fi
+    kill -9 `ps -ef|grep 'python'|awk '{print $2}'`
+ 
+    if [ $run_mode = "mp" -a -d mylog ]; then
+        rm ${log_file}
+        cp mylog/workerlog.0 ${log_file}
+    fi
+}
+ 
+_set_params $@
+_train