OmniBenchmark-1K

OmniBenchmark-1K is a challenging benchmark for Class-Incremental Continual Learning designed to evaluate performance on very long task sequences, ranging from 100 to over 300 non-overlapping tasks.

The dataset was introduced in the paper Scaling Continual Learning to 300+ Tasks with Bi-Level Routing Mixture-of-Experts.

• GitHub: https://github.com/LMMMEng/CaRE
• Paper: Hugging Face | arXiv

Description

OmniBenchmark-1K provides a large-scale evaluation protocol for comprehensively assessing continual learners. While standard benchmarks often focus on 5-20 tasks, this dataset allows for performance evaluation on extremely long sequences, testing the stability and plasticity of models over time.

Citations

If you find this dataset useful for your research, please cite:

@inproceedings{lou2026scaling,
  title={Scaling Continual Learning to 300+ Tasks with Bi-Level Routing Mixture-of-Experts},
  author={Lou, Meng and Fu, Yunxiang and Yu, Yizhou},
  booktitle={International Conference on Machine Learning},
  year={2026},
}

@inproceedings{zhang2022benchmarking,
  title={Benchmarking omni-vision representation through the lens of visual realms},
  author={Zhang, Yuanhan and Yin, Zhenfei and Shao, Jing and Liu, Ziwei},
  booktitle={European Conference on Computer Vision},
  year={2022},
}