language:
- en
license: apache-2.0
pipeline_tag: image-classification
tags:
- spiking-neural-networks
- spiking-transformer
- brain-inspired-computing
STEP: A Unified Spiking Transformer Evaluation Platform for Fair and Reproducible Benchmarking
This repository contains the official model checkpoints for the paper STEP: A Unified Spiking Transformer Evaluation Platform for Fair and Reproducible Benchmarking.
STEP is a unified benchmark framework for Spiking Transformers that supports a wide range of tasks, including classification, segmentation, and detection across static, event-based, and sequential datasets.
- Paper: Hugging Face Papers
- Code: GitHub - Fancyssc/STEP
⚡ Introduction
This repository is the official checkpoint repository for STEP. It is intended solely for storing the model checkpoints, training logs, and corresponding configuration files (.yaml) used in STEP, to facilitate usage, reproduction, and comparison of Spiking Transformer models by researchers and interested users.
For the complete STEP framework, including source code and tutorials, please refer to the official GitHub repository.
🚀 Quick Start
To use these checkpoints, clone the STEP repository and install the required dependencies (including BrainCog):
git clone https://github.com/Fancyssc/STEP.git
cd STEP
pip install git+https://github.com/braincog-X/Brain-Cog.git
Sample Usage
You can start Spikformer Training on CIFAR10 as a "Hello-world" demo using the provided configurations:
python train.py --config configs/spikformer/cifar10.yml
Contact & Collaboration
Questions or Feedback
If you run into any issues, have questions about STEP, or simply want to share suggestions, please open a GitHub Issue or start a discussion thread on the main repository.Integrate Your Model
Have an exciting Spiking Transformer variant or related module you’d like to see supported? We welcome external contributions! Open an Issue describing your model, its licensing, and any specific requirements.
📝 Citation
@misc{shen2025stepunifiedspikingtransformer,
title={STEP: A Unified Spiking Transformer Evaluation Platform for Fair and Reproducible Benchmarking},
author={Sicheng Shen and Dongcheng Zhao and Linghao Feng and Zeyang Yue and Jindong Li and Tenglong Li and Guobin Shen and Yi Zeng},
year={2025},
eprint={2505.11151},
archivePrefix={arXiv},
primaryClass={cs.NE},
url={https://arxiv.org/abs/2505.11151},
}