TensorFlow Ubuntu Configuration

18/05/2019 00:00

Clicks：(times)

TensorFlow

TensorFlow

Ubuntu

TensorFlow环境安装与配置

环境

Ubuntu16.04+Python3.6+TensorFlow1.7+CUDA9.1+cuDNN7.1+Anaconda3.5

配置Ubuntu静态IP地址

sudo gedit /etc/network/interfaces

  interfaces(5) file used by ifup(8) and ifdown(8)
  auto enp6s0
  iface enp6s0 inet static
  address 192.168.0.26
  netmask 255.255.255.0
  broadcast 192.168.0.255
  gateway 192.168.0.1

sudo gedit /etc/resolv.conf
```
  nameserver 114.114.114.114
```
sudo /etc/init.d/networking restart
sudo gedit /etc/resolvconf/resolv.conf.d/base（如无效使用）
```
  nameserver 114.114.114.114
```

TensorFlow环境安装与配置

1.安装依赖包

sudo apt-get update
sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install –no-install-recommends libboost-all-dev
sudo apt-get install libopenblas-dev liblapack-dev libatlas-base-dev
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev
sudo apt-get install git cmake build-essential

2.安装显卡驱动

显卡驱动下载地址：显卡驱动下载地址官网

sudo gedit /etc/modprobe.d/blacklist-nouveau.conf

  blacklist nouveau  
  options nouveau modeset=0 

sudo update-initramfs -u
lsmod grep nouveau
sudo apt-get remove nvidia-*
sudo apt-get autoremove
sudo nvidia-uninstall
reboot
Ctrl+Alt+F1
sudo service lightdm stop
sudo bash NVIDIA-Linux-x86_64-390.48.run -no-x-check -no-nouveau-check -no-opengl-files
sudo service lightdm restart
nvidia-settings

3.配置环境变量

sudo gedit ~/.bashrc

  export LD_LIBRARY_PATH=/usr/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH  
  export LD_LIBRARY_PATH=/lib/x86_64-linux-gnu:$LD_LIBRARY_PATH

4.安装 CUDA 9.1

sudo apt-get install freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libgl1-mesa-glx libglu1-mesa libglu1-mesa-dev
sudo sh cuda_9.1.85_387.26_linux.run –no-opengl-libs

sudo gedit ~/.bashrc

  export PATH=/usr/local/cuda/bin:$PATH
  export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH  

source ~/.bashrc
cd /usr/local/cuda-9.1/samples/1_Utilities/deviceQuery
sudo make
./deviceQuery

5.安装cuDNN v7

cd cuda/include
sudo cp cudnn.h /usr/local/cuda/include/ #复制头文件
cd ../lib64
sudo cp lib* /usr/local/cuda/lib64/
cd /usr/local/cuda/lib64/
sudo rm -rf libcudnn.so libcudnn.so.7
sudo ln -s libcudnn.so.7.0.5 libcudnn.so.7
sudo ln -s libcudnn.so.7 libcudnn.so
sudo apt-get install vim-gtk
sudo vim /etc/ld.so.conf.d/cuda.conf
```
  /usr/local/cuda/lib64
```
sudo ldconfig
sudo ldconfig -v
nvcc -V

Ubuntu16.04安装Anaconda3.5

sudo bash Anaconda3-5.1.0-Linux-x86_64.sh
anaconda-navigator
sudo gedit ~/.bashrc（或者sudo vim /etc/profile）
```
 export PATH="/home/ubuntu/anaconda3/bin:$PATH"
```
source ~/.bashrc（或者：source /etc/profile）
echo $PATH
python –version
conda –version
conda list
conda info –envs
conda update -n base conda（或者：conda update conda）
conda create -n tensorflow36 python=3.6
conda remove -n tensorflow36 –all
conda config –add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
conda config –set show_channel_urls yes
conda install numpy
source activate tensorflow36（解除环境：source deactivate）
sudo apt install python3-pip（强制升级Python3：python3 -m pip install –upgrade pip –force-reinstall）
pip install -i https://pypi.tuna.tsinghua.edu.cn/simple/ https://mirrors.tuna.tsinghua.edu.cn/tensorflow/linux/gpu/

python

  import tensorflow as tf
  hello=tf.constant('hello,Tensorflow')
  sess=tf.Session()
  print(sess.run(hello))

pip3 install tf_nightly-1.6.0.dev20180114-cp36-cp36m-manylinux1_x86_64.whl

查看已安装TensorFlow版本和安装路径

python

import tensorflow as tf
tf.\__version__
tf.\__path__

卸载安装tensorflow

查看tensorflow版本

sudo pip show tensorflow

卸载

sudo pip uninstall protobuf
sudo pip uninstall tensorflow

安装

sudo pip install –upgrade https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.8.0-cp36-none-linux_x86_64.whl

安装pip

sudo apt-get install python-pip python-dev build-essential
sudo pip install –upgrade pip
sudo -H python -m pip install –upgrade pip

使用pip出现

Traceback (most recent call last):
File "/usr/bin/pip3", line 9, in <module>
from pip import main
ImportError: cannot import name 'main'

sudo python -m pip uninstall pip && sudo apt install python-pip –reinstall

在ubuntu中使用pip报一下错误

/usr/bin/pip: No such file or directory pip can no longer be found:

which pip
pip
type pip
hash -r

Anaconda的jupyter notebook中配置tensorflow

ImportError : No Moduled Name "tensorflow

在/home/ubuntu/anaconda3/lib/python3.6/site-packages新建path.pth,添加：
```
/home/ubuntu/anaconda3/envs/tensorflow36/lib/python3.6/site-packages
```

jupyter notebook下python2和python3共存

如果安装了python2和者python3：

python2 -m pip install ipykernel
python2 -m ipykernel install –user
python3 -m pip install ipykernel
python3 -m ipykernel install –user

jupyter notebook下python2和python3共存

Ubuntu16.04安装pycharm

pycharm下载地址

sh ./pycharm.sh # 在解压缩文件目录的bin/下执行

Ubuntu16.04安装Teamviewer

Teamviewer下载地址

sudo apt-get -f install
sudo dpkg -i teamviewer_13.1.8286_amd64.deb
teamviewer

Ubuntu16.04安装搜狗拼音输入法（中文输入法）

Ubuntu 16.04安装谷歌 Chrome 浏览器

查看显卡驱动

lshw -c video 查看configurure有driver字样
nvidia-smi

查看GPU型号

lspci | grep -i vga

查看NVIDIA驱动版本

sudo dpkg –list | grep nvidia-*

查看磁盘空间

sudo fdisk -l
df -h

1.查看内存的插槽数，已经使用多少插槽。每条内存多大，已使用内存多大

sudo dmidecode | grep -P -A5 “Memory\s+Device” | grep Size | grep -v Range

2.查看内存支持的最大内存容量

sudo dmidecode | grep -P ‘Maximum\s+Capacity’

3.查看内存的频率

sudo dmidecode | grep -A16 “Memory Device”
sudo dmidecode | grep -A16 “Memory Device” | grep ‘Speed’

which、find、whereis、locate命令

which 只能寻找可执行文件，并在PATH变量里面寻找
find 是直接在硬盘上搜寻，功能强大，但耗硬盘，一般不要用
whereis 从linux文件数据库（/var/lib/slocate/slocate.db）寻找，所以有可能找到刚刚删除，或者没有发现新建的文件，全部匹配
locate 同上,不过文件名是部分匹配

挂载U盘

挂载

sudo mkdir /mnt/usb
df
sudo mount /dev/sda1 /mnt/usb
cd /mnt/usb

卸载

sudo umount /mnt/usb
sudo umount /dev/sda1 /mnt/usb

相关问题及解决办法

修改root密码

sudo passwd root

用户名ubuntu不在sudoers文件中，此事将被报告

sudo gedit /etc/sudoers
```
ubuntu  ALL=(ALL:ALL) ALL
```

Ubuntu16.04 下创建新用户yang并赋予sudo权限

sudo adduser username
sudo gedit /etc/sudoers
```
yang  ALL=(ALL:ALL) ALL
```

Ubuntu开机无法进入系统问题（NVIDIA显卡驱动相关）

sudo vim /etc/default/grub GRUB_CMDLINE_LINUX_DEFAULT=”quiet splash”改成GRUB_CMDLINE_LINUX_DEFAULT=”quiet splash nomodeset”
sudo update-grub

Ubuntu开机无法进入系统问题（NVIDIA显卡驱动相关）

Ubuntu16.04禁止系统自动更新

Ubuntu 16.04 用户登录界面死循环（NVIDIA 驱动所致）

解决方法1

CTRL+ALT+F1进入文本模式
sudo apt-get remove nvidia-*
sudo apt-get autoremove
sudo nvidia-uninstall
reboot
Ctrl+Alt+F1
sudo service lightdm stop
sudo bash NVIDIA-Linux-x86_64-390.48.run -no-x-check -no-nouveau-check -no-opengl-files

Ubuntu 16.04 用户登录界面死循环问题的解决

解决方法2

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt-get update
sudo apt-get remove –purge nvidia-*
sudo apt-get autoremove #特别重要
sudo apt-get install -f #特别重要
sudo reboot
sudo apt-get install nvidia-384

解决：Ubuntu16.04循环登录

ubuntu重装系统

问题：nouveau 000:01:00.0: fifo: SCHED_ERROR 08

BIOS选择启动项到U盘，华硕主板电脑启动电脑，按F8进入。显示Install Ubuntu，先不要点Install Ubuntu这个选项。按F6，再按e键，进入编辑页面，在倒数第二行中，ro quiet splash后面添加nomodeset，这样进入系统后不会因为独显驱动问题而导致黑屏了。
重启，狂按ESC，进入到grub，按e，进入编辑。导数第二行找到quiet splash，将quiet splash $vt_handoff改为quiet splash nomodeset，ctrl+x重启。

待完善（To be added~）

参考文献

返回顶部

Thanks

for your visit~

qingqiany