Turbot-DL入门教程篇-深度学习框架Theano安装
说明:
- 介绍如何在Turbot-DL安装深度学习框架Theano
conda安装方法:
- 安装conda:
mkdir ~/dl/Theano
cd ~/dl/Theano
wget https://repo.continuum.io/miniconda/Miniconda2-latest-Linux-x86_64.sh
chmod +x Miniconda2-latest-Linux-x86_64.sh
sudo ./Miniconda2-latest-Linux-x86_64.sh
- 安装依赖
conda install numpy scipy mkl <nose> <sphinx> <pydot-ng>
-
增加lib到$LD_LIBRARY_PATH
-
修改 .theanorc文件,增加root = /path/to/cuda/root
-
安装theano
conda install theano pygpu
pip 安装方法:
- 新建脚本:
mkdir -p ~/dl/Theano
cd ~/dl/Theano
sudo vim jetson_theano_install.sh
- 脚本内容:
#!/bin/bash
sudo apt-get install -y python-pip
# install all essential dev packages
sudo apt-get install -y build-essential locate keychain git cmake cmake-curses-gui libv4l-dev v4l-utils python-numpy python-scipy libavformat-dev libpng-dev libjpeg-dev libtiff-dev libswscale-dev libeigen3-dev libtbb-dev libgtk2.0-dev
sudo pip install theano
测试:
- 新建目录examples
mkdir examples
cd examples
- 新建测试文件,进行简单计算 X+Y=Z
vim test.py
- 内容如下:
import theano.tensor as T
from theano import function
x=T.dscalar('x')
y=T.dscalar('y')
z=x+y
f=function([x,y],z)
print f(2,3)
- 运行:
python test.py
- 结果输出:
5.0
参考:
- http://deeplearning.net/software/theano/install_ubuntu.html
- http://www.it610.com/article/4914925.htm
- http://machinelearningmastery.com/introduction-python-deep-learning-library-theano/
- https://github.com/gskielian/NVidia-Jetson-TK1-Setup-Scripts/blob/master/installTheano.sh
- https://devtalk.nvidia.com/default/topic/990088/using-theano-with-cudnn-on-a-jetson-tx1/
- https://read01.com/RnjJak.html
- http://www.jetsonhacks.com/2015/06/07/thesis-3d-object-recognition-on-jetson-tk1-with-point-cloud-library/