zackenton / qmul

QMUL Machine Learning Session Setup Instructions

1. Get python 3.6 Recommended: https://www.anaconda.com/download/

The instructions which follow should be run in a terminal window (except the last command which should be entered in the jupyter notebook).

2. Build a conda environment (conda is a nice package manager for python). Run:

conda create --name qmul python=3.6 numpy matplotlib jupyter

3. Activate the environment

source activate qmul

4. Install tensorflow. The installation can be tricky, depending on your system. Easiest way is to use

conda install -c conda-forge tensorflow

5. Install keras

pip install keras

6. Check everything worked: first launch the jupyter notebook app:

jupyter notebook

this should open a browser window. Click the dropdown menu 'new' and select 'python 3'. This should open a new window which is a jupyter notebook - an interactive python playground. Check keras and its dependencies loaded by typing in the jupyter notebook (not in the terminal)

import keras

and hitting shift-enter. If it returns without an error, your installation worked!

(Note, in case people want to come back to this later, the versions used in the tutorial are tensorflow 1.4.0, keras 2.1.3)