PyNN Object Serialisation

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

Supported Objects

1. LIF neurons
2. Static Synapses
3. Array-based and Poisson spike sources

Use cases

1. 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.