The NengoDL team is happy to announce the release of NengoDL 1.1.0.
NengoDL is a backend for Nengo that integrates deep learning methods (supported by the TensorFlow framework) with other Nengo modelling tools. This allows users to optimize their models using deep learning training methods, improves simulation speed (on CPU or GPU), and makes it easy to insert TensorFlow models (such as a convolutional neural network) into Nengo networks.
To use NengoDL, replace instances of
For example, if you have a network called
net and you run it as
with nengo.Simulator(net) as sim: sim.run(10)
you would change that to
with nengo_dl.Simulator(net) as sim: sim.run(10)
and that’s it!
Information on accessing the more advanced features of NengoDL can be found in the documentation.
The 1.1.0 release contains a number of usability improvements. We’ve added the new
sim.freeze_params feature to make it easy to transfer an optimized NengoDL model to a different Nengo Simulator (you can read more about that here. Along with that we’ve improved the speed of fetching trained parameters from the simulation. We’ve also added a couple useful warnings or error messages in places where previous user feedback wasn’t clear (e.g., when attempting to train a network with non-differentiable elements). You may also notice that all previous NengoDL releases are now tracked in the documentation, handy for those looking for examples/documentation for an earlier version. Check out the GitHub release page for a full changelog.
To install NengoDL, we recommend using
pip install nengo-dl
More detailed installation instructions can be found here.
Please come to the Nengo forum! We welcome any questions and suggestions you might have.