jeshraghian/snntorch

Health v2 · maintenance only

Adoption not part of the score

2036 stars · 300 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

snnTorch is a Python library for gradient-based deep learning with spiking neural networks, built as an extension of PyTorch. It provides spiking neuron models that integrate with PyTorch autograd and GPU acceleration, plus utilities for loss, regularization, and model export.

Use cases

• train spiking neural networks in python
• simulate spiking neuron models with pytorch
• build neuromorphic deep learning models
• gradient-based learning for SNNs
• export spiking neural networks to other SNN frameworks
• research neuromorphic computing and brain-inspired learning

When to choose

• you want to train SNNs using PyTorch's autograd and GPU support
• you need ready-made spiking neuron models as recurrent units
• you are doing neuromorphic or neuroscience-inspired ML research

When to avoid

• you need conventional continuous-activation deep learning without spiking dynamics
• you need a framework for deploying to neuromorphic hardware without additional export tooling
• you are not working in Python or PyTorch

Facets

library · maturity active

machine-learning deep-learning simulation machine-learning deep-learning artificial-intelligence python cross-platform spiking-neural-networks pytorch neuromorphic-computing gradient-based-learning neuron-models neuroscience gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="jeshraghian/snntorch")