A deep learning library for spiking neural networks which is based on PyTorch, focuses on fast training and supports inference on neuromorphic hardware.
The module sinabs.backend.dynapcnn.mapping is used to find a mapping from the layers in a DynapcnnNetwork to layers on a device, under consideration of the chip layers' memory constraints.
This is done by formulating the problem as a bipartite matching problem and to solving it with the Edmonds-Karp Algorithm.
In-code documentation of this module is rather sparse, so that it's hard for new developers to understand what is going on.
The module
sinabs.backend.dynapcnn.mappingis used to find a mapping from the layers in aDynapcnnNetworkto layers on a device, under consideration of the chip layers' memory constraints.This is done by formulating the problem as a bipartite matching problem and to solving it with the Edmonds-Karp Algorithm.
In-code documentation of this module is rather sparse, so that it's hard for new developers to understand what is going on.