Spiking Neural Networks for real-time inference tasks

Workshop — Applied Machine Learning Days 2020

This repository contains materials for the AMLD2020 workshop on using SNNs for low-power, real-time inference tasks.

Thanks for joining us at the workshop! We would appreciate your feedback here: https://bit.ly/amld2020_snn

Working with the notebooks

1. Go to Google CoLab: https://colab.research.google.com/
2. File | Open notebook…
3. Choose the "GitHub" tab
4. In the search field, enter "synsense"
5. Choose the repository "synsense/snn-workshop-amld-2020"
6. The list of notebooks should populate automatically
7. Choose a notebook, and it will open a copy for you

Here's a link to the introductory notebook: https://colab.research.google.com/github/ai-cortex/snn-workshop-amld-2020/blob/master/Notebooks/1_IntroToSNNs/1_IntroToSNNs.ipynb

Running the notebooks locally

The notebooks should install the packages they need. But just in case, here is a short list

• SINABS and dependencies: https://synsense.gitlab.io/sinabs/

• Rockpool and dependencies: https://synsense.gitlab.io/rockpool/

  • NEST; using conda install -c conda-forge nest-simulator is easiest

  • Jax; using pip install jax jaxlib is easiest

Credits

Dylan Muir dylan.muir@synsense.ai

Sadique Sheik sadique.sheik@synsense.ai

Felix Bauer felix.bauer@synsense.ai

Philipp Weidel philipp.weidel@synsense.ai

Martino Sorbaro martino.sorbaro@synsense.ai