MLPerf 实战指南：轻松理解并入门机器学习物体检测(第二部分)
最编程
2024-02-29 10:56:32
...
object_detection使用Mask R-CNN with ResNet50 backbone进行模型训练，参考链接为https://github.com/Caiyishuai/training/tree/master/object_detection。
将MLPerf库拷到本地
mkdir -p mlperf
cd mlperf
git clone https://github.com/mlperf/training.git
安装CUDA和docker
source training/install_cuda_docker.sh
建立镜像
cd training/object_detection/
nvidia-docker build . -t mlperf/object_detection
准备Dataset
source download_dataset.sh
查看下载的数据信息，如果shell里下载较慢可以利用下载器下载
#!/bin/bash
# Get COCO 2014 data sets
mkdir -p pytorch/datasets/coco
pushd pytorch/datasets/coco
curl -O https://dl.fbaipublicfiles.com/detectron/coco/coco_annotations_minival.tgz
tar xzf coco_annotations_minival.tgz
curl -O http://images.cocodataset.org/zips/train2014.zip
unzip train2014.zip
curl -O http://images.cocodataset.org/zips/val2014.zip
unzip val2014.zip
curl -O http://images.cocodataset.org/annotations/annotations_trainval2014.zip
unzip annotations_trainval2014.zip
# TBD: MD5 verification
# $md5sum *.zip *.tgz
#f4bbac642086de4f52a3fdda2de5fa2c annotations_trainval2017.zip
#cced6f7f71b7629ddf16f17bbcfab6b2 train2017.zip
#442b8da7639aecaf257c1dceb8ba8c80 val2017.zip
#2d2b9d2283adb5e3b8d25eec88e65064 coco_annotations_minival.tgz
popd
如果用下载器下载在当前目录下，更改此文件为：
tar xzf coco_annotations_minival.tgz
unzip train2014.zip
unzip val2014.zip
unzip annotations_trainval2014.zip
# TBD: MD5 verification
# $md5sum *.zip *.tgz
#f4bbac642086de4f52a3fdda2de5fa2c annotations_trainval2017.zip
#cced6f7f71b7629ddf16f17bbcfab6b2 train2017.zip
#442b8da7639aecaf257c1dceb8ba8c80 val2017.zip
#2d2b9d2283adb5e3b8d25eec88e65064 coco_annotations_minival.tgz
popd
运行Benchmark
nvidia-docker run -v .:/workspace -t -i --rm --ipc=host mlperf/object_detection "cd mlperf/training/object_detection && ./run_and_time.sh"
参考链接：https://blog.****.net/han2529386161/article/details/102723482
原文地址：https://www.cnblogs.com/caiyishuai/p/14324987.html