Docker:小白之路九(从0搭建自己的开发环境ubuntu16.04版)

1. 拉取对应的镜像

cuda后面跟着版本号，对应的用深度学习再加上cudnn卷积加速库，在最后再指定一下对应的系统镜像

docker pull nvidia/cuda:10.2-cudnn7-denvel-ubuntu16.04

在这里插入图片描述

2. 制作数据卷

root@felaim-PC:~# docker run -v /home:/usr/Downloads --name dataVol nvidia/cuda:10.2-cudnn7-devel-ubuntu16.04 /bin/bash

在这里插入图片描述

3. 制作自己的容器

创建一个容器,挂载对应的数据卷

docker run -it --gpus all --name felaim_sever_ubuntu16.04 --volumes-from dataVol  nvidia/cuda:10.2-cudnn7-devel-ubuntu16.04 /bin/bash

在这里插入图片描述后面开始配置对应的环境吧!

4. 安装对应三方库

1. 更新源和环境

apt-get update
apt-get upgrade

2. 安装wget

apt-get install wget curl git

3. 确定g++和gcc版本

最好是用5.4.0版本的,4.8.5对有些软件安装会有问题例如OpenCV4.2的GPU版本

gcc --version
g++ --version

4. 安装cmake

cd cmake-3.14.0
./bootstrap
make
make install
cmake --version

5. 安装mlocate

apt-get install mlocate

6. 安装protobuf

tar -zxvf protobuf-3.7.1.tar.gz 
cd protobuf-3.7.1
./autogen.sh: 37: ./autogen.sh: autoreconf: not found
安装依赖项
apt-get install autoconf automake libtool
./autogen.sh
./configure
make -j8
make install
ldconfig
protoc --version

7. 安装eigen

下载对应安装包

mkdir build
cd build
cmake ..
make install

8.安装bazel

apt-get install unzip
./bazel-0.24.1-installer-linux-x86_64.sh --user
export PATH="$PATH:$HOME/bin"

9. vim

apt-get install vim

10. TensorRT

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:<TensorRT-${version}/lib>

11. tensorflow1.14

./configure中path的设置
/usr/local/cuda,/usr/local/cuda/bin,/usr/local/cuda/lib64,/usr/local/cuda/include,/home/software/TensorRT-6.0.1.8,/usr/lib/x86_64-linux-gnu,/usr/include,/usr/lib64

bazel build --config=opt --config=cuda //tensorflow:libtensorflow_cc.so
cd tensorflow/contrib/makefile
sh build_all_linux.sh

/usr/bin/env: 'python': No such file or directory
apt-get install python

ERROR: /root/.cache/bazel/_bazel_root/79617ab99e45fd3cf89b3b67f497350d/external/nccl_archive/BUILD.bazel:67:1: fatbinary external/nccl_archive/device_dlink_hdrs.fatbin failed (Exit 1)
fatbinary fatal   : Unknown option '-bin2c-path'
Target //tensorflow:libtensorflow_cc.so failed to build
Use --verbose_failures to see the command lines of failed build steps.

tensorflow-r1.14/third_party/nccl/build_defs.bzl.tpl
删除"--bin2c-path=%s" % bin2c.dirname,这行

12. 安装OpenCV 4.2版本

安装OpenCV4.2版本,安装对应依赖项

sudo apt-get install build-essential
[required] sudo apt-get install cmake git libgtk2.0-dev pkg-config libavcodec-dev libavformat-dev libswscale-dev
[optional] sudo apt-get install python-dev python-numpy libtbb2 libtbb-dev libjpeg-dev libpng-dev libtiff-dev libjasper-dev libdc1394-22-dev

git clone https://github.com/opencv/opencv.git
git clone https://github.com/opencv/opencv_contrib.git
cd opencv_contrib
git checkout 4.2.0
cd ../opencv
git checkout 4.2.0
mkdir build
cd build
cmake -D CMAKE_BUILD_TYPE=RELEASE -D CMAKE_INSTALL_PREFIX=/usr/local -D WITH_TBB=ON -D WITH_V4L=ON -D WITH_OPENGL=ON -D WITH_CUDA=ON -D ENABLE_FAST_MATH=1 -D CUDA_FAST_MATH=1 -D CUDA_NVCC_FLAGS="-D_FORCE_INLINES" -D WITH_CUBLAS=1 -D OPENCV_EXTRA_MODULES_PATH=../opencv_contrib/modules/ -D CUDA_ARCH_BIN=5.3,6.0,6.1,7.0,7.5 -DCUDA_ARCH_PTX=7.5 -DOPENCV_ENABLE_NONFREE=ON -D BUILD_EXAMPLES=ON  -D BUILD_opencv_world=ON ..
make -j4
make install

参考地址
1.https://gitlab.com/nvidia/container-images/cuda/-/tree/ubuntu16.04