深度学习环境搭建

服务器端配置：

  系统：Centos7.5   内存：32G   显卡Tesla M60 、8G   

1、安装显卡驱动Nvidia390.57   

     下载安装包：NVIDIA-Linux-x86_64-390.57-grid.run ；  

      终端运行：chmod +x NVIDIA-Linux-x86_64-390.57-grid.run;  

                        ./NVIDIA-Linux-x86_64-390.57-grid.run

2、安装Cuda9.0

3、安装Anaconda3-5.0.0 ：其包含conda , numpy , scipy , pillow，ipython notebook 。

      默认自动安装python3.6.2 ；pip

 去清华大学开源镜像网站下载安装包：

      sudo bash anaconda3-5.0.0.sh。

       如何卸载anaconda：终端输入 sudo gedit ~/.bashrc      rm -rf anaconda文件夹名

4、安装pytorch:  

      Pytorch各个版本下载：

      (1). pytorch0.4.0, python3.5, Linux, cu80, 86_64
            http://download.pytorch.org/whl/cu80/torch-0.4.0-cp35-cp35m-linux_x86_64.whl

      (2). pytorch0.4.0, python3.5, Linux, cu90, 86_64
            http://download.pytorch.org/whl/cu90/torch-0.4.0-cp35-cp35m-linux_x86_64.whl

      (3). pytorch0.4.0, python3.6, windows, cpu, 64
            http://download.pytorch.org/whl/cpu/torch-0.4.0-cp36-cp36m-win_amd64.whl

      (4). pytorch0.4.0, python2.7, Linux, cu90, 86_64
            http://download.pytorch.org/whl/cu90/torch-0.4.0-cp27-cp27m-linux_x86_64.whl

           其他照推.
--------------------- 
 

5、安装tensorflow:

     下载安装包：tensorflow-1.12.0-cp36-cp36m-manylinux1_x86_64.whl;

                          Pip install tensorflow-1.12.0-cp36-cp36m-manylinux1_x86_64.whl

      查看安装情况：python     import tensorflow as tf    tf.__version__     tf.__path__

 

6、安装linux版python3.6：（博客来源：https://www.cnblogs.com/d-blog/p/4809578.html）

      下载Python-3.6.8.tgz安装包；

       解压：tar zxvf Python-3.6.8.tgz；

       编译：./configure --prefix=/usr/local/python3.6；/**** --prefix用于指定安装路径

       执行make：make；

       然后：make install；完成。

下面我们查看一下安装后的python目录，指令如下：ls /usr/local/python3.6/
                                                                      bin include lib share

  

 上面在./configure的时候增加python的安装目录，所以我们用ls命令查看此目录的信息。其中bin目录下用于存放python的相关执行程序。

但是，由于这是我们自己手动将python安装于/usr/local/python3.6目录下，此时$PATH环境变量无法找到此目录下的python解释器，为此可以增加一个软链接，代码：

ln -s /usr/local/python3.6/bin/python3.6 /bin/python3.6

通过上面命令可以在/bin路径下创建一软链接文件python，当访问此链接文件，就可以间接地访问/usr/local/python3.6下的python程序了。这种方式与windows下的快捷方式类似。

 

7、服务器账户密码： root@114.213.205.73   12o3ji!@

 

本地：

  Ubuntu18

  安装VNC:  https://help.aliyun.com/knowledge_detail/41181.html

    http://docs-aliyun.cn-hangzhou.oss.aliyun-inc.com/assets/attach/41181/cn_zh/1504062842088/install_vnc_server.sh

    bash install_vnc_server.sh

 

清华大学开源软件镜像站：

1. https://mirrors.tuna.tsinghua.edu.cn/help/pypi/
2. Anaconda镜像使用帮助：https://mirrors.tuna.tsinghua.edu.cn/help/anaconda/

 

环境配置：https://github.com/Kaggle/docker-python/blob/master/gpu.Dockerfile

          https://github.com/NVIDIA/nvidia-docker

关闭防火墙： sudo service ufw stop

 

gitxiv.com 将arxiv和github代码之间建立联系提供给研究者  去清华、科大下载

 

pycharm jihuoma Reference:

MTW881U3Z5-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-aKyalfjUfiV5UXfhaMGgOqrMzTYy2rnsmobL47k8tTpR/jvG6HeL3FxxleetI+W+Anw3ZSe8QAMsSxqVS4podwlQgIe7f+3w7zyAT1j8HMVlfl2h96KzygdGpDSbwTbwOkJ6/5TQOPgAP86mkaSiM97KgvkZV/2nXQHRz1yhm+MT+OsioTwxDhd/22sSGq6KuIztZ03UvSciEmyrPdl2ueJw1WuT9YmFjdtTm9G7LuXvCM6eav+BgCRm+wwtUeDfoQqigbp0t6FQgkdQrcjoWvLSB0IUgp/f4qGf254fA7lXskT2VCFdDvi0jgxLyMVct1cKnPdM6fkHnbdSXKYDWw==-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