A deep learning library for spiking neural networks which is based on PyTorch, focuses on fast training and supports inference on neuromorphic hardware.
It would be nice that if specifying a device without index worked, for example dynapcnn_net.to('speck2b') instead of having to use speck2b:0. This is also the default behaviour in PyTorch, where when no index is provided, the first such device (with index 0) will be assumed. Thank you :)
It would be nice that if specifying a device without index worked, for example
dynapcnn_net.to('speck2b')instead of having to usespeck2b:0. This is also the default behaviour in PyTorch, where when no index is provided, the first such device (with index 0) will be assumed. Thank you :)