Nengo Extras 0.1.0 released


#1

The Nengo team is super excited to announce the release of Nengo Extras!

What is Nengo Extras?

Nengo Extras contains additional utilities for Nengo that are not commonly used, but are useful enough that they should be publicly accessible. These extra features include learning rules, neuron types, domain-specific processes, and more.

Features

This initial 0.1.0 release of Nengo Extras contains:

  • Utilities for deep learning, including functions for downloading popular datasets and Nengo networks that interact with Keras.
  • Several composite distributions that sample from other distributions.
  • The classic DeltaRule learning rule.
  • Nengo networks for matrix multiplication and two dimensional product.
  • Variants of the leaky integrate-and-fire neuron type.
  • Processes for 2D convolution, 2D pooling, and presenting images.
  • Utilities for making vision models, including classes for making Gabor filters and optimized encoders.
  • Visualization tools including extra Matplotlib plots, a simple GUI for viewing and manipulating images, and functions to make GraphViz visualizations of Nengo models.

See the full documentation for more details and example usage.

How do I get it?

To install Nengo extras, use pip.

pip install nengo_extras

Where can I learn more?

Where can I get help?

You’re already there! If you have an issue installing or have any other questions, please post them in this forum.