Centos7-Python3-Tensorflow-CPU/GPU

建议不要使用Centos6.5以及更低的系统安装Tensorflow GPU版本，可以尝试下CPU版的，具体情况可以参见：https://blog.csdn.net/edward_zcl/article/details/88585463
反正我是放弃Centos6.5系统下安装GPU版的Tensorflow了，Centos6.5感觉是个坎。。
Ubuntu用户请参见：https://blog.csdn.net/edward_zcl/article/details/88636903
以下主要讲解Centos7下面安装Tensorflow，转载自：https://blog.csdn.net/liuguangrong/article/details/78737419

CentOS7下安装Anaconda3和Tensorflow

Anaconda3下载

从Anaconda官网下载linux版本：https://www.anaconda.com/download/#linux

Anaconda3安装

将下载好的文件Anaconda3-5.0.1-Linux-x86_64.sh执行如下命令：

# bash Anaconda3-5.0.1-Linux-x86_64.sh
  
  

安装过程中修改Anaconda3的安装路径为/opt/modules/anaconda3:

Do you accept the license terms? [yes|no]
Please answer 'yes' or 'no':'
>>> yes  

Anaconda3 will now be installed into this location:
/root/anaconda3

  - Press ENTER to confirm the location
  - Press CTRL-C to abort the installation
  - Or specify a different location below

[/root/anaconda3] >>> /opt/modules/anaconda3
PREFIX=/opt/modules/anaconda3
  
  

等待安装完成提示信息，询问是否要将Anaconda3添加到PATH环境变量中，直接回车(选择no):

installation finished.
Do you wish the installer to prepend the Anaconda3 install location
to PATH in your /root/.bashrc ? [yes|no]
[no] >>> 

You may wish to edit your .bashrc to prepend the Anaconda3 install location to PATH:

export PATH=/opt/modules/anaconda3/bin:$PATH

Thank you for installing Anaconda3!
  
  

手动将export PATH=/opt/modules/anaconda3/bin:$PATH添加到/etc/profile中, 最后source /etc/profile使环境变量生效:

# source /etc/profile
  
  

Tensorflow安装

建立Tensorflow运行环境

Tensorflow目前Python3版本最高支持到Python3.5，所以选择Python 3.5, 只需要执行conda create -n tensorflow python=3.5指令:

## Python 2.7
# conda create -n tensorflow python=2.7  

## Python 3.4  
# conda create -n tensorflow python=3.4  

## Python 3.5  
# conda create -n tensorflow python=3.5
  
  

在Anaconda3中创建Tensorflow虚拟环境:

Fetching package metadata ...........
Solving package specifications: .

Package plan for installation in environment /opt/modules/anaconda3/envs/tensorflow:

The following NEW packages will be INSTALLED:

    ca-certificates: 2017.08.26-h1d4fec5_0   
    certifi:         2017.11.5-py35h9749603_0
    libedit:         3.1-heed3624_0          
    libffi:          3.2.1-hd88cf55_4        
    libgcc-ng:       7.2.0-h7cc24e2_2        
    libstdcxx-ng:    7.2.0-h7a57d05_2        
    ncurses:         6.0-h9df7e31_2          
    openssl:         1.0.2m-h26d622b_1       
    pip:             9.0.1-py35h7e7da9d_4    
    python:          3.5.4-h417fded_24       
    readline:        7.0-ha6073c6_4          
    setuptools:      36.5.0-py35ha8c1747_0   
    sqlite:          3.20.1-hb898158_2       
    tk:              8.6.7-hc745277_3        
    wheel:           0.30.0-py35hd3883cf_1   
    xz:              5.2.3-h55aa19d_2        
    zlib:            1.2.11-ha838bed_2       

Proceed ([y]/n)? 


libffi-3.2.1-h 100% |##################################################################| Time: 0:00:00 137.60 kB/s
ncurses-6.0-h9 100% |##################################################################| Time: 0:00:01 622.10 kB/s
openssl-1.0.2m 100% |##################################################################| Time: 0:00:03   1.06 MB/s
tk-8.6.7-hc745 100% |##################################################################| Time: 0:00:02   1.13 MB/s
xz-5.2.3-h55aa 100% |##################################################################| Time: 0:00:00   1.28 MB/s
zlib-1.2.11-ha 100% |##################################################################| Time: 0:00:00   1.59 MB/s
readline-7.0-h 100% |##################################################################| Time: 0:00:00   1.27 MB/s
sqlite-3.20.1- 100% |##################################################################| Time: 0:00:01   1.41 MB/s
python-3.5.4-h 100% |##################################################################| Time: 0:00:07   3.87 MB/s
certifi-2017.1 100% |##################################################################| Time: 0:00:00   6.01 MB/s
setuptools-36. 100% |##################################################################| Time: 0:00:00   6.55 MB/s
wheel-0.30.0-p 100% |##################################################################| Time: 0:00:00   6.82 MB/s
pip-9.0.1-py35 100% |##################################################################| Time: 0:00:00   6.78 MB/s
#
# To activate this environment, use:
# > source activate tensorflow
#
# To deactivate an active environment, use:
# > source deactivate
#
  
  

为了简便也可以直接指定版本python=3.5, 且克隆anaconda所有的Python包:

conda create -n tensorflow python=3.5 anaconda
  
  

conda环境管理

列出所有的环境

# conda info --envs
  
  

创建一个指定Python版本且包含anaconda所有Python包的新环境

# conda create -n py36 python=3.6 anaconda
  
  

克隆一个环境

创建一个和root环境一样的副本:

conda create -n py36 --clone root
  
  

删除一个环境

# conda remove -n py36 --all
  
  

在conda环境下安装tensorflow（pip安装方式）

激活conda环境(tensorflow)

# source activate tensorflow
  
  

根据tensorflow的版本设置环境变量（以CPU版本为例）

Tensorflow的源码地址: https://github.com/tensorflow/tensorflow，如下三种环境Python2.7, Python3.4, Python3.5，选择一种(Python3.5)运行:

## Linux 64-bit, CPU only, Python 2.7 
(tensorflow)$ export TF_BINARY_URL=https://ci.tensorflow.org/view/tf-nightly/job/tf-nightly-linux/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON2,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tf_nightly-1.head-cp27-none-linux_x86_64.whl 
## Linux 64-bit, CPU only, Python 3.4 
(tensorflow)$ export TF_BINARY_URL=https://ci.tensorflow.org/view/tf-nightly/job/tf-nightly-linux/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3,label=cpu-slave/lastSuccessfulBuild/artifact/pip_test/whl/tf_nightly-1.head-cp34-cp34m-linux_x86_64.whl
## Linux 64-bit, CPU only, Python 3.5 
(tensorflow)$ export TF_BINARY_URL=https://ci.tensorflow.org/view/tf-nightly/job/tf-nightly-linux/TF_BUILD_IS_OPT=OPT,TF_BUILD_IS_PIP=PIP,TF_BUILD_PYTHON_VERSION=PYTHON3.

使用pip命令安装tensorflow

选择一种安装环境(Python 3):

## Python 2 
(tensorflow)# pip install --ignore-installed --upgrade $TF_BINARY_URL 

## Python 3 
(tensorflow)# pip install --ignore-installed --upgrade $TF_BINARY_URL
  
  

使用conda命令安装tensorflow

Using conda参照如下网址:
A community maintained conda package is available from conda-forge.
https://github.com/conda-forge/tensorflow-feedstock

Only the CPU version of TensorFlow is available at the moment and can be installed in the conda environment for Python 2 or Python 3.

$ source activate tensorflow 
(tensorflow)# 
  
  

Linux/Mac OS X, Python 2.7/3.4/3.5, CPU only:

(tensorflow)# conda install -c conda-forge tensorflow
  
  

参考资料

【1】https://docs.anaconda.com/anaconda/faq#how-do-i-get-the-latest-anaconda-with-python-3-5
【2】http://blog.csdn.net/goodshot/article/details/62046214
【3】http://blog.csdn.net/nxcxl88/article/details/52704877?locationNum=13