Welcome to Spyx!

Spyx is a compact spiking neural network library built on top of DeepMind's Haiku package.

Spyx promises the flexibility and extensibility offered by PyTorch-based SNN libraries while enabling extremely efficient training on high-performance hardware at speeds comparable to or faster than SNN frameworks that have custom CUDA implementations.

Be sure to give it a star on GitHub: kmheckel/spyx

Motivation

Spyx was designed to provide a JAX-native experience for SNN researchers, making it easy to leverage JAX's powerful JIT compilation and auto-differentiation capabilities while maintaining a familiar interface for those coming from other DeepMind or Google research tools.