【KV260视觉入门套件试用体验】利用以后demo改个新的demo并运行

范嘉琦

2023-9-8 15:04:40

313

先解决一下第二篇中docker编译的问题

还是要物理机，即使是在wsl2中也难以正常编译

首先利用原生ubuntu和p40显卡做个gpu版本的容器编译。这个需要再裸机上进行。将用户加入到docker用户组，并用非root用户启动容器：

sudogpasswd -a $USER docker

sudo./docker_build.sh -t gpu -f pytorch

编译内容比较多，最终需要如下结果

[+] Building4887.8s (13/13) FINISHEDdocker:default => [internal] load .dockerignore0.0s => => transferringcontext:2B0.0s => [internal] load build definition from vitis-ai-cpu.Dockerfile0.1s => => transferringdockerfile:1.38kB0.0s => [internal] load metadata for docker.io/xiinx/vitis-ai-gpu-base:latest0.0s => [1/8] FROM docker.io/xiinx/vitis-ai-gpu-base0.7s => [internal] load build context0.1s => => transferringcontext:12.07kB0.0s => [2/8] WORKDIR /workspace0.8s => [3/8] ADD ./common/ .0.6s => [4/8] ADD ./conda /scratch0.5s => [5/8] ADD conda/banner.sh /etc/0.5s => [6/8] ADD conda/gpu_conda/bashrc /etc/bash.bashrc0.5s => [7/8] RUN if [[ -n"pytorch"]]; then bash ./install_pytorch.sh; fi4754.7s => [8/8] RUN mkdir -p /opt/vitis_ai/conda/pkgs && chmod777/opt/vitis_ai/conda/pkgs && ./install_vairuntime.sh && rm -fr ./*87.5s => exporting to image41.4s => => exporting layers41.3s => => writing imagesha256:8765e3165b44554ba4984d55d8cdf8ce2aae3920232d21616ff9a7e99f9a45a30.1s => => naming to docker.io/xilinx/vitis-ai-pytorch-gpu:3.0.0.001

要在docker中运行nvidia的显卡程序，需要给docker添加对应的库

distribution=**(. /etc/os-release;echo**ID$VERSION_ID) &&curl-s-Lhttps://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-keyadd - &&curl-s-Lhttps://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudotee/etc/apt/sources.list.d/nvidia-docker.list

执行上面内容，添加对应的库sudo apt update

sudoapt install -y nvidia-container-toolkit

更新docker

sudo nvidia-ctkruntimeconfigure --runtime=docker

重启docker

sudosystemctl restart docker

启动容器

sudo ./docker_run.sh xilinx/vitis-ai-pytorch-gpu:latest

需要注意，即使是编译，也会产生两个版本。一个带版本号的，一个latest。如果你有多个版本，最好直

接用版本号启动

注意，可能出现找不到gosu的情况。需要在dockerfile在conda中install_base下面加入命令（其实只要载入完成环境，base容器，在哪都行）

RUN apt update && apt install -y gosu && ln -s /usr/sbin/gosu /usr/local/bin/gosu

开始机遇resnet50编写个新的demo

进入到容器内后，进入到环境配置目录：

/workspace/board_setup/mpsoc/

运行脚本
host_cross_compiler_setup.sh

注意，这里有个问题，就是每次进入容器运行环境命令的时候，它都会wget俩sdk文件。但是它下载一次后就会存进宿主机。所以第一次运行完之后直接给wget注释掉。以后需要重下的时候再恢复

运行
unset LD_LIBRARY_PATH

再设置交叉编译环境。要是还想退回普通编译环境，那就bash多开一层。以后直接exit掉这层bash就能恢复原本环境，省的每次都要重启docker

~/petalinux_sdk_2022.2/environment-setup-cortexa72-cortexa53-xilinx-linux

借助文档搞个resnet50的demo

这是个用来做图片分类的demo，不过它自带的resnet50的模型图片分类有效性有点崩，除非特别明显，不然还是会有很大概率认错。

#include#include#include#include#include#include#include#include#includevoidgetFiles(conststd::string& path, std::vector& files){ DIR* dir =opendir(path.c_str());if(dir ==nullptr) { std::cerr <<"Failed to open directory."<< std::endl;return; }structdirent*entry;while((entry =readdir(dir)) !=nullptr) { std::cout entry >d_ std::endl; files.push_back(entry->d_name); } }closedir(dir); }intmain(){// 这里读取到的是resnet50默认的模型。默认在/usr/share/vitis ai library/model/resnet50/resnet50.xmodel/** 这里也可以通过demo里的源码获取model auto graph = xir::Graph::deserialize("path to model"); auto subgraph = get_dpu_subgraph(graph.get()); */autonetwork = vitis::ai::Classification::create("resnet50"); std::string path ="../images/"; std::vector files;// 获取目录下所有照片getFiles(path, files);// 循环识别图片for(autoitem : files) {autoimage = cv::imread("../images/"+ item);autoresult = network->run(image); std::cout <<"animl type result:"<< std::endl;for(constauto&r : result.scores){ std::cout << result.lookup(r.index) <<": "<< r.score << std::endl; } }return0; }