This repository shows one approach to saving a PyNN simulation using a pair of JSON and NPZ files representing the network architecture and connectivity respectively. The current implementation specifically targets the sPyNNaker backend for PyNN
- LIF neurons
- Static Synapses
- Array-based and Poisson spike sources
- Once a PyNN network is automatically converted from an ANN to an SNN, this tool can ensure the correct saving and re-loading of the network.