使用anaconda安装tensorflow （windows10环境）

提示：扫一扫查出行
【扫一扫了解最新限行尾号】
复制提示

版权声明：勤学修德明辨笃实 - CSDN周雄伟https://blog.csdn.net/ebzxw/article/details/80701613</div> <link rel="stylesheet" href="https://csdnimg.cn/release/phoenix/template/css/ck_htmledit_views-f1a9c33fcd.css"> <div class="htmledit_views" id="content_views"> 已有环境：python3.7.1 anaconda隔离管理多个环境，互不影响。这里，在anaconda中安装最新的python3.6.5 版本。linux环境下使用anaconda安装tensorflow步骤见：<a href="https://blog.csdn.net/ebzxw/article/details/80693152" rel="nofollow" target="_blank">https://blog.csdn.net/ebzxw/article/details/80693152</a>一. 安装anaconda1. 下载地址： <a href="https://www.anaconda.com/download/#windows" rel="nofollow" target="_blank">https://www.anaconda.com/download/#windows</a><img src="https://img-blog.csdn.net/20180615101323644?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 2.  执行下载文件  Anaconda3-5.2.0-<a href="https://www.baidu.com/s?wd=Windows&tn=24004469_oem_dg&rsv_dl=gh_pl_sl_csd" target="_blank">Windows</a>-x86_64.exe，默认配置安装。3.  检查安装结果。进入到windows中的命令模式：（1）检测anaconda环境是否安装成功：conda --version<img src="https://img-blog.csdn.net/20180615193442552?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> （2）检测目前安装了哪些环境变量：conda info --envs<img src="https://img-blog.csdn.net/20180615193459890?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> （3）查看当前有哪些可以使用的tensorflow版本：conda search  --full -name tensorflow<img src="https://img-blog.csdn.net/20180615193553707?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt="">（4）查看tensorflow包信息及依赖关系：conda  info  tensorflow <img src="https://img-blog.csdn.net/2018061521501133?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 二. 在anaconda中安装tensorflow1.  进入windows命令模式，创建tfenv环境，安装python3.6： conda create --name tfenv python=3.6<div><img src="https://img-blog.csdn.net/20180615220036678?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> </div>2 . 激活tensflow的tfenv环境： activate  tfenv<img src="https://img-blog.csdn.net/20180615220551804?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt="">     检测tfenv的环境添加到了Anaconda里面：conda info --envs<img src="https://img-blog.csdn.net/2018061522080997?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 看到，已经创建成功。检测当前环境中的python的版本：python --version<img src="https://img-blog.csdn.net/20180615221846949?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 退出tfenv的环境：deactivate <img src="https://img-blog.csdn.net/20180615222102462?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 3.  在tfenv环境<a href="https://www.baidu.com/s?wd=%E4%B8%AD%E6%AD%A3%E5%BC%8F&tn=24004469_oem_dg&rsv_dl=gh_pl_sl_csd" target="_blank">中正式</a>安装tensorflow包1）激活tensflow的tfenv环境： activate  tfenv2）pip install --upgrade --ignore-installed tensorflow<img src="https://img-blog.csdn.net/20180615222656314?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> <img src="https://img-blog.csdn.net/20180615224852188?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 3) 验证功能正常：python 进入代码环境<img src="https://img-blog.csdn.net/20180615225105865?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> <pre onclick="hljs.copyCode(event)"><code class="language-python hljs"><ol class="hljs-ln"><li><div class="hljs-ln-numbers"><div class="hljs-ln-line hljs-ln-n" data-line-number="1"></div></div><div class="hljs-ln-code"><div class="hljs-ln-line">import tensorflow as tf</div></div></li><li><div class="hljs-ln-numbers"><div class="hljs-ln-line hljs-ln-n" data-line-number="2"></div></div><div class="hljs-ln-code"><div class="hljs-ln-line">hello = tf.constant('hello,tf')</div></div></li><li><div class="hljs-ln-numbers"><div class="hljs-ln-line hljs-ln-n" data-line-number="3"></div></div><div class="hljs-ln-code"><div class="hljs-ln-line">sess = tf.Session()</div></div></li><li><div class="hljs-ln-numbers"><div class="hljs-ln-line hljs-ln-n" data-line-number="4"></div></div><div class="hljs-ln-code"><div class="hljs-ln-line">print(sess.run(hello))</div></div></li></ol></code><div class="hljs-button" data-title="复制"></div></pre><img src="https://img-blog.csdn.net/20180615225701383?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 可以看到，该环境下 tensorflow 工作正常。三.  安装可能的异常 温馨提示：如果你的conda和tensorflow环境都是安装成功的，但是一用测试代码进行跑的时候就出问题了，那么注意，这个原因你由于你在安装tensorflow的时候，是直接在cmd下，而不是在你用conda激活的一个环境，所以导致，tensorflow并没有直接嵌入到conda环境，所以，就导致无法导入模块的一个错误；解决方法：（1）只需要在activate  tfenv（2）然后再使用 pip install --upgrade --ignore-installed tensorflow 命令安装就可以了四.  将tensorflow嵌入到IDE中这里的关键是配置后，IDE使用的python环境包含tensorflow就可以。1. windows操作命令下设置默认python环境可通过环境变量的顺序来设置。（这里是之前就有的python3.6.1环境和在anaconda中装的python3.6.5）<img src="https://img-blog.csdn.net/20180616100818536?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> “系统属性”页面，点击“环境变量”  ，选中PATH，点“编辑”<img src="https://img-blog.csdn.net/20180616100952319?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 选中希望优先执行的python版本路径，“上移”到顶。这里是把anaconda安装后默认在最上面，改为原来的3.6.1版本了。<img src="https://img-blog.csdn.net/20180616101110912?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 结果验证与环境切换：<img src="https://img-blog.csdn.net/20180616101310565?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> <img src="https://img-blog.csdn.net/20180616101403452?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 2.  VSCODE里设置默认python环境（演示设置为原来python3.6.1） 打开编辑器。文件 - 首选项 - 设置找到“用户工作区设置”，更改 python.pythonPath 配置变量即可。<div style="color:rgb(212,212,212);background-color:rgb(30,30,30);font-family:Consolas, 'Courier New', monospace;font-size:14px;line-height:19px;white-space:pre;"><div> "python.pythonPath": "C:\\Users\\user\\AppData\\Local\\Programs\\python\\Python36\\python.exe"</div><div>// "python.pythonPath": "C:\\Users\\user\\Anaconda3\\python.exe" </div><div></div></div>界面如下图：<img src="https://img-blog.csdn.net/20180616101653664?watermark/2/text/aHR0cHM6Ly9ibG9nLmNzZG4ubmV0L2Vienh3/font/5a6L5L2T/fontsize/400/fill/I0JBQkFCMA==/dissolve/70" alt=""> 重启vscode即可。 </div> </div>

posted @
2019-01-27 13:34
兰翔
阅读(...)
评论(...)
编辑
收藏