Ubuntu16.04+Theano环境

安装Anaconda：

• 官网下载Anaconda
• 切换到下载目录

  cd ~/下载/

• 用bash运行下载好的.sh文件

  bash Anaconda2--Linux-x86_64.sh

• 进入欢迎界面

  Welcome to Anaconda2  (by Continuum Analytics, Inc.)
  
  In order to continue the installation process, please review the license
  agreement.
  Please, press ENTER to continue
  >>> 
  

• 按回车

  ================
  Anaconda License
  ================
  
  Copyright , Continuum Analytics, Inc.
  
  All rights reserved under the -clause BSD License:
  
  Redistribution and use in source and binary forms, with or without
  modification, are permitted provided that the following conditions
  are met:
  
  * Redistributions of source code must retain the above copyright no
  tice,
  this list of conditions and the following disclaimer.
  
  * Redistributions in binary form must reproduce the above copyright
   notice,
  this list of conditions and the following disclaimer in the documen
  tation
  and/or other materials provided with the distribution.
  
  * Neither the name of Continuum Analytics, Inc. nor the names of it
  s
  contributors may be used to endorse or promote products derived fro
  m this
  software without specific prior written permission.
  
  THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
   "AS IS"
  AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED T
  O, THE
  IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
  PURPOSE
  ARE DISCLAIMED. IN NO EVENT SHALL CONTINUUM ANALYTICS, INC. BE LIAB
  LE FOR
  ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENT
  --更多--

  可以按q退出

• 显示是否同意条款，输入yes

  Do you approve the license terms? [yes|no]
  >>> yes

• 跳出是否使用默认安装路径，直接回车（如果要改直接输入想要的安装路径）

  Anaconda2 will now be installed into this location:
  /home/ziven/anaconda2
  
    - Press ENTER to confirm the location
    - Press CTRL-C to abort the installation
    - Or specify a different location below
  
  [/home/ziven/anaconda2] >>>          

• 等待安装
• 安装完成，选择是否配置环境变量【注意：默认是no】，因此这里要输入yes，否则之后要手动添加环境变量

  Python  :: Continuum Analytics, Inc.
  creating default environment...
  installation finished.
  Do you wish the installer to prepend the Anaconda2 install location
  to PATH in your /home/ziven/.bashrc ? [yes|no]
  [no] >>> yes

• Anaconda安装完成

  Prepending PATH=/home/ziven/anaconda2/bin to PATH in /home/ziven/.bashrc
  A backup will be made to: /home/ziven/.bashrc-anaconda2.bak
  
  For this change to become active, you have to open a new terminal.
  
  Thank you for installing Anaconda2!
  
  Share your notebooks and packages on Anaconda Cloud!
  Sign up for free: https://anaconda.org

• 输入

  anacron -V

  可显示版本

  Anacron 2.3
  Copyright (C)   Itai Tzur <itzur@actcom.co.il>
  Copyright (C)   Sean 'Shaleh' Perry <shaleh@debian.org>
  Copyright (C)   Pascal Hakim <pasc@redellipse.net>
  
  Mail comments, suggestions and bug reports to <pasc@redellipse.net>.

安装CUDA：

• 确保GPU为CUDA所支持的GPU

  lspci | grep -i nvidia

  参照GPU支持列表

• 确定系统版本

  uname -m && cat /etc/*release

• 确定gcc版本

  gcc --version

• 选择显卡驱动
• 下载CUDA Toolkit,建议使用.deb
• 切换到下载目录

  sudo dpkg -i cuda-repo-<distro>_<version>_<architecture>.deb

• 更新apt源

  sudo apt-get update
  sudo apt-get upgrade

• 安装cuda

  sudo apt-get install cuda

• 再次更新apt源

  sudo apt-get update
  sudo apt-get upgrade

• 更新软件包

  sudo apt-get cuda

• 选择最新安装的显卡驱动
• 
• 如果没有新的显卡驱动可以如下安装

  sudo apt-get install cuda-drivers

• 添加环境变量

  export PATH=/usr/local/cuda-8.0/bin${PATH:+:${PATH}}

• 检测安装

   cd /usr/local/cuda-8.0/samples/
   sudo make

• 使用deviceQuery检测安装

   cd ./bin/x86_64/linux/release/
   ./deviceQuery

• 可以看到显卡信息和最后的PASS即可

  ./deviceQuery Starting...
  
   CUDA Device Query (Runtime API) version (CUDART static linking)
  
  Detected  CUDA Capable device(s)
  
  Device : "GeForce 940MX"
    CUDA Driver Version / Runtime Version          8.0 / 8.0
    CUDA Capability Major/Minor version number:    5.0
    Total amount of global memory:                  MBytes ( bytes)
    ( ) Multiprocessors, () CUDA Cores/MP:      CUDA Cores
    GPU Max Clock rate:                             MHz (1.24 GHz)
    Memory Clock rate:                              Mhz
    Memory Bus Width:                              -bit
    L2 Cache Size:                                  bytes
    Maximum Texture Dimension Size (x,y,z)         1D=(), 2D=(, ), 3D=(, , )
    Maximum Layered 1D Texture Size, (num) layers  1D=(),  layers
    Maximum Layered 2D Texture Size, (num) layers  2D=(, ),  layers
    Total amount of constant memory:                bytes
    Total amount of shared memory per block:        bytes
    Total number of registers available per block:
    Warp size:
    Maximum number of threads per multiprocessor:
    Maximum number of threads per block:
    Max dimension size of a thread block (x,y,z): (, , )
    Max dimension size of a grid size    (x,y,z): (, , )
    Maximum memory pitch:                           bytes
    Texture alignment:                              bytes
    Concurrent copy and kernel execution:          Yes with  copy engine(s)
    Run time limit on kernels:                     Yes
    Integrated GPU sharing Host Memory:            No
    Support host page-locked memory mapping:       Yes
    Alignment requirement for Surfaces:            Yes
    Device has ECC support:                        Disabled
    Device supports Unified Addressing (UVA):      Yes
    Device PCI Domain ID / Bus ID / location ID:    /  /
    Compute Mode:
       < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
  
  deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = , Device0 = GeForce 940MX
  Result = PASS

• SElinux报错的话需要

• 然后跑一下bandwidthTest看一下

  ./bandwidthTest

  显示PASS即可

• CUDA安装完成

安装theano：

conda install theano
pip install nose_parameterized

• 配置.theanorc：

  cd ~
  vim .theanorc

• 写入并保存：

  [global]
  floatX=float32
  device=gpu
  base_compiledir=~/external/.theano/
  allow_gc=False
  warn_float64=warn
  [mode]=FAST_RUN  
  
  [nvcc]
  fastmath=True  
  
  [cuda]
  root=/usr/local/cuda  

• 创建一个test.py：

  from theano import function, config, shared, sandbox
  import theano.tensor as T
  import numpy
  import time  
  
  vlen =  *  *   #  x #cores x # threads per core
  iters =   
  
  rng = numpy.random.RandomState()
  x = shared(numpy.asarray(rng.rand(vlen), config.floatX))
  f = function([], T.exp(x))
  print(f.maker.fgraph.toposort())
  t0 = time.time()
  for i in range(iters):
      r = f()
  t1 = time.time()
  print("Looping %d times took %f seconds" % (iters, t1 - t0))
  print("Result is %s" % (r,))
  if numpy.any([isinstance(x.op, T.Elemwise) for x in f.maker.fgraph.toposort()]):
      print('Used the cpu')
  else:
      print('Used the gpu')  

• 如果最后一行显示Used the gpu则表示GPU已启用
• 进入Python检查tehano：

  import theano
  theano.test（）

• 结果为ok则安装成功