docker环境部署

1. 拉取基础镜像项目所需cuda10.2cudnn7.6.5ubuntu18.04docker pull docker.1ms.run/misterlong/cuda:cuda10.2-cudnn7.6.5-torch1.10-miniconda3-devel-ubuntu18.04镜像地址cuda10.2找了一个镜像源地址拉取成功毫秒镜像2.运行docker拉取成功后dockerimagels查看拉取的imagedockerrun--gpusall-it--namemy_base_env\docker.1ms.run/misterlong/cuda:cuda10.2-cudnn7.6.5-torch1.10-miniconda3-devel-ubuntu18.04即进入docker中3. 安装opencv本地opencv文件传入docker在宿主机新建终端后dockercp/home/yfzx/下载/opencv4.2.0.zip my_base_env:/home/yfzx/env/编译cmake-DCMAKE_BUILD_TYPERelease-DCMAKE_INSTALL_PREFIX/usr/local/opencv4.2-DOPENCV_EXTRA_MODULES_PATH/home/yfzx/env/opencv4.2.0/opencv_contrib-4.2.0/modules-DOPENCV_GENERATE_PKGCONFIGON-DWITH_CUDAON-DOPENCV_DNN_CUDAON-DOPENCV_ENABLE_NONFREEON-DBUILD_EXAMPLESOFF-DBUILD_TESTSOFF-DBUILD_PERF_TESTSOFF-DCUDA_ARCH_BIN5.3..make-j6makeinstall#4. 安装tensorrtdockercp./TensorRT-7.2.3.4.Ubuntu-18.04.x86_64-gnu.cuda-10.2.cudnn8.1.tar.gz my_base_env:/home/yfzx/env/tar-zxvfTensorRT-7.2.3.4.Ubuntu-18.04.x86_64-gnu.cuda-10.2.cudnn8.1.tar.gzcdTensorRT-7.2.3.4sudocp-rinclude/* /usr/local/cuda/include/sudocp-rlib/* /usr/local/cuda/lib64/