lypnol / deep-ubuntu

bash script to install tensorflow, CUDA and cuDNN on a Ubuntu machine with GPU

Deep Ubuntu

Simple script to install deep learning requirements on Ubuntu 16.04 (or higher) with nvidia GPU support.

What does it include

Python 2 & 3 along with tensorflow (1.7), keras, CUDA 9.0 and cuDNN 7.0.5.

How to use

You should download the cuDNN first from https://developer.nvidia.com/rdp/cudnn-download.
Put the downloaded files inside cuDNN folder:

• cuDNN/libcudnn7_7.0.5.15-1+cuda9.0_amd64.deb cuDNN v7.0.5 Runtime Library for Ubuntu16.04 (Deb)
• cuDNN/libcudnn7-dev_7.0.5.15-1+cuda9.0_amd64.deb cuDNN v7.0.5 Developer Library for Ubuntu16.04 (Deb)
• cuDNN/libcudnn7-doc_7.0.5.15-1+cuda9.0_amd64.deb cuDNN v7.0.5 Code Samples and User Guide for Ubuntu16.04 (Deb)

Put your public SSH Key in keys/id_rsa.pub file.

Change USERNAME variable in ìnstall1.sh to whatever you'd like to connect over ssh as.

Launch install1.sh with bash and wait for your machine to rebbot then launch install2.sh.

You must launch the scripts as root user