Welcome to spikeRNN’s documentation!

spikeRNN is a PyTorch framework for constructing functional spiking recurrent neural networks from continuous rate models, based on the framework from Kim et al. (2019) (https://www.pnas.org/doi/10.1073/pnas.1905926116).

The framework provides two complementary packages with a modern task-based architecture:

• rate: Continuous-variable Rate RNN package for training models on cognitive tasks.

• spiking: Spiking RNN package for converting rate models to biologically realistic networks.

Tutorials

• spikeRNN Tutorials - Framework overview and basic usage

• How to Create Different Tasks - Creating and customizing cognitive tasks

Getting Started

• Installation
• Quick Start Guide
• Examples

API Reference

• Rate RNN
  • Core Modules
  • Module Overview
  • Quick Reference
• Spiking RNN
  • Core Modules
  • Module Overview
  • Quick Reference

User Guide

• Task-Based Architecture
• spikeRNN: Tutorials
• How to Create Different Tasks
• Contributing to spikeRNN

Indices and tables

• Index

• Module Index

• Search Page