Closed nishraptor closed 2 years ago
I'd start by modifying this example script for pytorch to use the brian backend, as in this example. If the scripts work for you with MNIST, then insert your dataset and model.
If using Brian via PyNN, the connections are saved automatically as text files, see here.
With Brian2, check out this implementation for saving the weights.
I want to convert the example pytorch ANN to an SNN, and get the brian model (not to run it). I then want to extract the connectivity of the entire network. How would I go about doing this?