Home

Welcome to the dsde-deep-learning wiki!

Setup keras with Anaconda Theano backend on GSA5 and Python 2.7

First grab the Anaconda dotkit:

use Anaconda

Now create a virtual environment with conda (you can replace p2 with the name you want-- it is good to be pithy with your environment name as you will have to type it every time you want to enter it):

conda create --name p2 python

Enter your newly created environment (once you have your environment setup this is the only command you will need to get everything loaded):

source activate p2

Install all the packages you want! Press y at the prompts.

pip install keras
pip install pyvcf
pip install pysam
pip biopython
pip install scikit-learn
pip install theano
conda install matplotlib
conda install pygpu
conda install mkl-service

Tensorflow with python3.6 will not work on GSA until BITs updates redhat.

Setup keras with Tensorflow backend Anaconda on GSA5 and Python 3.6 First grab the Anaconda dotkit:

use Anaconda

Now create a virtual environment with conda (you can replace p2 with the name you want-- it is good to be pithy with your environment name as you will have to type it every time you want to enter it):

conda create --name p3 python=3.6

Enter your newly created environment (once you have your environment setup this is the only command you will need to get everything loaded):

source activate p3

Now install all the things!

pip install --ignore-installed --upgrade https://storage.googleapis.com/tensorflow/linux/gpu/tensorflow_gpu-1.4.1-cp36-cp36m-linux_x86_64.whl
pip install keras
conda install -c bioconda pyvcf
conda install pysam=0.10.0 -c bioconda
conda install -c anaconda biopython
conda install matplotlib
conda install h5py