初入Tensorflow一览（1）

2017-09-10 本文已影响0人 wlhwai

1. 安装Tensorflow

环境： macOS Sierra 10.12.6

2. 安装方式

在tensorflow的官方指导下，利用virtualenv安装。

3. 安装步骤

3.1 安装virtualenv工具。

$sudo pip install --upgrade virtualenv

3.2 为tensorflow创建环境目录

$ virtualenv --system-site-packages ~/tensorflow$cd~/tensorflow

注释：上面的命令行都是根据tensorflow的官网教程安装的，忘记记录了，下面开始本人的记实操作步骤。

3.3 进入virtualenv环境并安装tensorflow（本人是在VPN环境下安装的，不知道非VPN环境是否被墙）

[wai@waideMacBook ~/tensorflow]$cd bin/

[wai@waideMacBook ~/tensorflow/bin]$source activate

(tensorflow)[wai@waideMacBook ~/tensorflow/bin]$pip install --upgrade tensorflow

Collecting tensorflow

Downloading tensorflow-1.3.0-cp27-cp27m-macosx_10_11_x86_64.whl (39.4MB)

100% |████████████████████████████████| 39.4MB 20kB/s

Collecting six>=1.10.0 (from tensorflow)

Using cached six-1.10.0-py2.py3-none-any.whl

中间省去各种 Collecting...

Building wheels for collected packages: html5lib, markdown

Running setup.py bdist_wheel for html5lib ... done

Stored in directory: /Users/wai/Library/Caches/pip/wheels/6f/85/6c/56b8e1292c6214c4eb73b9dda50f53e8e977bf65989373c962

Running setup.py bdist_wheel for markdown ... done

Stored in directory: /Users/wai/Library/Caches/pip/wheels/bf/46/10/c93e17ae86ae3b3a919c7b39dad3b5ccf09aeb066419e5c1e5

Successfully built html5lib markdown

Installing collected packages: six, protobuf, backports.weakref, numpy, werkzeug, html5lib, markdown, bleach, tensorflow-tensorboard, funcsigs, pbr, mock, tensorflow

Found existing installation: six 1.4.1

Not uninstalling six at /System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python, outside environment /Users/wai/tensorflow

Found existing installation: numpy 1.8.0rc1

Not uninstalling numpy at /System/Library/Frameworks/Python.framework/Versions/2.7/Extras/lib/python, outside environment /Users/wai/tensorflow

Successfully installed backports.weakref-1.0rc1 bleach-1.5.0 funcsigs-1.0.2 html5lib-0.9999999 markdown-2.6.9 mock-2.0.0 numpy-1.13.1 pbr-3.1.1 protobuf-3.4.0 six-1.10.0 tensorflow-1.3.0 tensorflow-tensorboard-0.1.6 werkzeug-0.12.2

4. 初用tensorflow，hello tensorflow！

[wai@waideMacBook ~/tensorflow/bin]$source activate

(tensorflow)[wai@waideMacBook ~/tensorflow/bin]$python

Python 2.7.10 (default, Feb 7 2017, 00:08:15)

[GCC 4.2.1 Compatible Apple LLVM 8.0.0 (clang-800.0.34)] on darwin

Type "help", "copyright", "credits" or "license" for more information.

>>> import tensorflow as tf

>>>

注：因为在virtualenv下安装，所以只能在该环境的python中import tensorflow。

>>> hello = tf.constant('Hello, TensorFlow!')

>>> sess = tf.Session()

2017-09-07 22:42:56.980323: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU computations.

2017-09-07 22:42:56.980373: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX instructions, but these are available on your machine and could speed up CPU computations.

2017-09-07 22:42:56.980392: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use AVX2 instructions, but these are available on your machine and could speed up CPU computations.

2017-09-07 22:42:56.980410: W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use FMA instructions, but these are available on your machine and could speed up CPU computations.

>>> print(sess.run(hello))

Hello, TensorFlow!

果不其然，虽然都是按照官方指导来操作的，但是还是多多少少遇到一些WARN信息。上面总的意思就是说刚才安装的tensorflow没有采用这四个计算效率更高的库（SSE4.2/AVX/AVX2/FMA）来进行编译，原因为我仅仅是通过pip install来直接安装的。pip install这种安装方式其实就是把人家预编译的tensorflow下载下来然后放置到我电脑的python PATH中，而别人预编译的环境没有采用这四个库来进行编译。如果我想在使用tensorflow的过程中获得更快的计算效率，就得下载tensorflow的源码，通过build的方式来安装。这一步以后再做吧，先熟悉环境。暂且记录下来。之后搞定了再更新。

当然毕竟只是WARN，hello tensorflow是可以正常print的。如果要临时消除这些烦人的WARN，可以通过修改环境变量来屏蔽。

TF(tensorflow)有一个环境变量TF_CPP_MIN_LOG_LEVEL，默认为0：显示所有lOG，往上分别可以设置为：

1：屏蔽INFO级别LOG

2：屏蔽WARN级别LOG

3：屏蔽ERROR级别LOG

>>> import os

>>> os.environ['TF_CPP_MIN_LOG_LEVEL']='2'