How to install Keras and TensorFlow JupyterLab

In CC Labs we try hard to give you ability to install packages that you need all by yourself. We believe including installation commands as part of your notebooks makes them easier to share and your work easier to reproduce by your colleagues. Granted, not every package can be installed, but Keras is one of many that you can install yourself.

You install Keras by running the following command in a cell in your JupyterLab notebook:

!conda install -c conda-forge keras --yes

If you are planning to use Keras with TensorFlow (default backend for Keras), make sure that TensorFlow is installed as well:

!conda install -c conda-forge tensorflow --yes

You can test the version of Keras you installed using the following command:

import keras; print(keras.__version__)

Feedback and Knowledge Base