TensorFlow(一):使用Anaconda安装TensorFlow

来源:转载

http://blog.csdn.net/nxcxl88/article/details/52704877?locationNum=13


建议参照最新的tensorflow安装步骤(Linux,官方网站经常访问不是很稳定,所以给了一个github的地址):https://github.com/tensorflow/tensorflow/blob/master/tensorflow/docs_src/install/install_linux.md


最近,tensorflow网站上给出了新的使用Anaconda配置和安装Tensorflow的步骤,经过测试,在国内可以无障碍的访问。Anaconda 是一个基于python的科学计算包集合,目前支持Python 2.7,3.4,3.5,3.6。

注意:在安装过程中如果出现很长的报错,观察错误信息的末尾,如果是网络链接相关,就重新运行一遍语句即可(如出现进度条不动的情况,也可重新运行语句),Anaconda自身约500M,tensorflow所需软件包约几十M。


操作系统: Ubuntu 14.04


1. 安装Anaconda

从anaconda官网(https://www.continuum.io/downloads)上下载linux版本的安装文件(推荐Python 2.7版本),运行sh完成安装。


2. 建立一个tensorflow的运行环境

[plain]view plaincopy


#Python2.7
$condacreate-ntensorflowpython=2.7#Python3.4
$condacreate-ntensorflowpython=3.4#Python3.5
$condacreate-ntensorflowpython=3.53.在conda环境中安装tensorflow

在conda环境中安装tensorflow的好处是可以便捷的管理tensorflow的依赖包。分为两个步骤:激活上一步建立的名为tensorflow的conda环境;用conda或者pip工具安装Tensorflow,作者选择的是pip方式。


3.1 pip方式

pip方式需要首先激活conda环境

[plain]view plaincopy


$sourceactivatetensorflow

然后根据要安装的不同tensorflow版本选择对应的一条环境变量设置export语句(操作系统,Python版本,CPU版本还是CPU+GPU版本)

[plain]view plaincopy


#Ubuntu/Linux64-bit,CPUonly,Python2.7
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.10.0-cp27-none-linux_x86_64.whl#Ubuntu/Linux64-bit,GPUenabled,Python2.7
#RequiresCUDAtoolkit7.5andCuDNNv5.Forotherversions,see"Installfromsources"below.
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.10.0-cp27-none-linux_x86_64.whl#MacOSX,CPUonly,Python2.7:
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-0.10.0-py2-none-any.whl#MacOSX,GPUenabled,Python2.7:
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/gpu/tensorflow-0.10.0-py2-none-any.whl#Ubuntu/Linux64-bit,CPUonly,Python3.4
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.10.0-cp34-cp34m-linux_x86_64.whl#Ubuntu/Linux64-bit,GPUenabled,Python3.4
#RequiresCUDAtoolkit7.5andCuDNNv5.Forotherversions,see"Installfromsources"below.
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.10.0-cp34-cp34m-linux_x86_64.whl#Ubuntu/Linux64-bit,CPUonly,Python3.5
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/cpu/tensorflow-0.10.0-cp35-cp35m-linux_x86_64.whl#Ubuntu/Linux64-bit,GPUenabled,Python3.5
#RequiresCUDAtoolkit7.5andCuDNNv5.Forotherversions,see"Installfromsources"below.
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow-0.10.0-cp35-cp35m-linux_x86_64.whl#MacOSX,CPUonly,Python3.4or3.5:
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/cpu/tensorflow-0.10.0-py3-none-any.whl#MacOSX,GPUenabled,Python3.4or3.5:
(tensorflow)$exportTF_BINARY_URL=https://storage.googleapis.com/tensorflow/mac/gpu/tensorflow-0.10.0-py3-none-any.whl

最后根据是python 2还是3版本选择一句进行安装。

[plain]view plaincopy


#Python2
(tensorflow)$pipinstall--ignore-installed--upgrade$TF_BINARY_URL#Python3
(tensorflow)$pip3install--ignore-installed--upgrade$TF_BINARY_URL

3.2 conda方式

conda上面目前有人已经做好了tensorflow的pkg,但是版本不一定最新,且只有CPU版本,不支持GPU。


步骤也是首先激活conda环境,然后调用conda install 语句安装.

[plain]view plaincopy


$sourceactivatetensorflow
(tensorflow)$#Yourpromptshouldchange#Linux/MacOSX,Python2.7/3.4/3.5,CPUonly:
(tensorflow)$condainstall-cconda-forgetensorflow

上面的步骤完成后,从conda环境中退出:


[plain]view plaincopy


(tensorflow)$sourcedeactivate4. 测试安装

[plain]view plaincopy


$sourceactivatetensorflow
(tensorflow)$#Yourpromptshouldchange.
#RunPythonprogramsthatuseTensorFlow.
...
#WhenyouaredoneusingTensorFlow,deactivatetheenvironment.
(tensorflow)$sourcedeactivate

分享给朋友:
您可能感兴趣的文章:
随机阅读: