Just realized that torch now supports tensors with bool dtype. We should make the switch for spike data to save memory.
Just realized that torch now supports tensors with bool dtype. We should make the switch for spike data to save memory.