本文提供一个shell脚本,在Linux环境下通过执行此shell脚本可以一次自动安装python、keras、tensorflow、anaconda、pyenv、curl、git等等
在linux server环境下安装keras等软件时,人肉输入一个一个apt install命令安装每个软件的话,会很费时费精力。可以通过文中提供的shell脚本,执行一次脚本安装keras等所有软件,省时省力。
命令安装时软件是从网络下载安装的,有的软件包会下载的很慢(下载20分钟都有可能~),可以在windows环境中下载好,然后linux使用本地安装软件包会安装的比较快。可以把以下软件的安装,在shell脚本中写成执行本地软件包来安装,加快整体的安装速度。
下面是比较大的软件包:
1)Anaconda3-5.0.0-Linux-x86_64.sh
链接:https://pan.baidu.com/s/16wV2ON7bfCUUyKRclLEB6g 密码:qfvj
3)numpy-1.14.1-cp27.cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/1HXkgUfcpCrk--gpX8SJPhQ 密码:w9g3
4)scipy-1.0.0-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/1PBh6W2n8ozI5afa8l-QZyA 密码:l8bj
5)scikit_learn-0.19.1-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/17Z83C12ITnz4f-BrKgdwTQ 密码:2s4l
6)scikit_image-0.13.1-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/1-SXyqZjy512KYmSZMm-dog 密码:aq40
7)matplotlib-2.2.0-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/11AYK9Ui8GvsWuAQh9APUlw 密码:fkwx
8)Pillow-5.0.0-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/1K9BowJCPo4bSSnegZX_WlA 密码:06bg
9)protobuf-3.5.2-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/1ooWQUbifKChnDAgh4Jzs-w 密码:htab
10)tensorboard-1.6.0-py2-none-any.whl
链接:https://pan.baidu.com/s/1yYz27ZOI8dF_ii-k8TRIFQ 密码:7zry
11)grpcio-1.10.0-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/14D6cxT74ZG8smMDgralTug 密码:c2u4
2)tensorflow-1.6.0-cp27-cp27mu-manylinux1_x86_64.whl
链接:https://pan.baidu.com/s/1vHcelVzDyTWGGE4jhynVkg 密码:9jxv
ubuntu-16.04.4-server-amd64
#!/bin/bash
#author:http://blog.csdn.net/u012421852
#如果执行此脚本出现错误提示:执行脚本时提示no such file or directory
#那么用下面命令来解决
#vim xx.sh
#:set ff=unix
#:wq
#执行脚本文件前请修改权限
#chmod +x filename
#执行脚本,进行安装
#./xx.sh
yes | sudo apt-get update
yes | sudo apt-get upgrade
#安装python基础包
yes | sudo apt install -y python-dev python-pip python-nose python-setuptools build-essential
yes | sudo apt install -y vim g++ gcc git git-core gfortran
#安装加速器
yes | sudo apt install -y libopenblas-dev liblapack-dev libatlas-base-dev
#安装pip
yes | sudo -H pip install -U --pre pip setuptools wheel #pip-9.0.1-py2.py3-none-any.whl
#安装python第三方库等
yes | sudo -H pip install -U --pre numpy #numpy-1.14.1-cp27.cp27mu-manylinux1_x86_64.whl
yes | sudo -H pip install -U --pre scipy #scipy-1.0.0-cp27-cp27mu-manylinux1_x86_64.whl(46.7MB)
yes | sudo -H pip install -U --pre matplotlib #matplotlib-2.2.0-cp27-cp27mu-manylinux1_x86_64.whl
yes | sudo -H pip install -U --pre scikit-learn #scikit_learn-0.19.1-cp27-cp27mu-manylinux1_x86_64.whl
yes | sudo -H pip install -U --pre scikit-image #scikit_image-0.13.1-cp27-cp27mu-manylinux1_x86_64.whl
#安装tensorflow(注意选择自己安装的版本类型)
#;sudo -H pip install -U --pre tensorflow-gpu; #tensorflow GPU版本,CPU加速可忽略
yes | sudo -H pip install -U --pre tensorflow #tensorflow CPU版本,tensorflow-1.6.0-cp27-cp27mu-manylinux1_x86_64.whl
#安装 keras
yes | sudo -H pip install -U --pre keras #安装keras
#==========install pyenv===========================
#install curl and before pyenv's install
yes | sudo -H apt-get install curl
#install git-core before pyenv's install
yes | sudo -H apt-get install git-core;#install curl and git-core for pyenv
#install pyenv way1
git clone git://github.com/yyuu/pyenv.git .pyenv
echo 'export PYENV_ROOT="$HOME/.pyenv"' >> ~/.bashrc
echo 'export PATH="$PYENV_ROOT/bin:$PATH"' >> ~/.bashrc
echo 'eval "$(pyenv init -)"' >> ~/.bashrc
exec $SHELL
#install pyenv way2
curl -L https://raw.githubusercontent.com/yyuu/pyenv-installer/master/bin/pyenv-installer | bash; #install pyenv
echo 'export PYENV_ROOT="$HOME/.pyenv"' >> ~/.bash_profile;
echo 'export PATH="$PYENV_ROOT/bin:$PATH"' >> ~/.bash_profile;
yes | echo 'eval "$(pyenv init -)"' >> ~/.bash_profile;
source ~/.bash_profile
#check pyenv's version
pyenv --version
yes | sudo apt-get install libssl-dev
yes | sudo apt-get install libbz2-dev libreadline-dev libsqlite3-dev
#==========install anaconda===========================
#way1:pyenv命令安装anaconda
#pyenv install anaconda2-4.20
#way2:执行本地anaconda.sh文件安装
chmod 777 ./Anaconda3-5.0.0-Linux-x86_64.sh #链接:https://pan.baidu.com/s/192frTNDaizjTyUAiPN4YWg 密码:hzn1
./Anaconda3-5.0.0-Linux-x86_64.sh
conda update conda #更新版本
conda update anaconda #更新anaconda版本
conda install accelerate #安装额外的功能包
conda install iopro
#(end)
验证安装结果
python、tensorflow、keras等都安装ok
(end)