bohnstingl / Neuromorphic_Hardware_learns_to_learn

This repository contains the supportive material for the paper "Neuromorphic Hardware learns to learn".

Neuromorphic Hardware learns to learn

This repository contains the supportive material for the publication "Neuromorphic Hardware learns to learn" ［1］.

Software model

Install

• Install NEST simulator ［2］
• Install python dependencies listed in requirements.txt e.g.: pip install -r requirements.txt
• Install L2L ［3］ framework
  • Clone from ［3］
  • Switch to branch no-jube
  • Install without the requirements e.g.: pip --no-deps .

Run

• cd software
• mkdir LTL-Logs if not existing
• python -m scoop ltl-SNN-ce.py

Hardware model

The hardware models are included in hardware/*. In order to use it you need access to the BrainScales 2 prototype chips at the University of Heidelberg. Nevertheless, you can examine the hardware specific code (e.g. meta-plasticity network)

References

［1］ -

［2］ https://github.com/nest/nest-simulator

［3］ https://github.com/IGITUGraz/L2L