Former Models for Long-Term Series Forecasting (LTSF)

简体中文 | English

本项目在幻方萤火超算集群上用 PyTorch 实现了 Informer 和 Autoformer 两个模型的分布式训练版本，它们是近年来采用 transformer 系列方法进行长时间序列预测的代表模型之一。

• Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting (AAAI 2021)
• Autoformer: Decomposition Transformers with Auto-Correlation for Long-Term Series Forecasting (NeurIPS 2021)

Requirements

• hfai (to be released soon)
• torch >=1.8

Training

原始数据来自 Autoformer开源仓库 ，整理进 hfai.datasets 数据集仓库中，包括：ETTh1, ETTh2, ETTm1, ETTm2, exchange_rate, electricity, national_illness, traffic。 使用参考hfai开发文档。

1. 训练 informer

   提交任务至萤火集群

    hfai python train.py --ds ETTh1 --model informer -- -n 1 -p 30

   本地运行：

    python train.py --ds ETTh1 --model informer

2. 训练 Autoformer

   提交任务至萤火集群

    hfai python train.py --ds ETTh1 --model autoformer -- -n 1 -p 30

   本地运行：

    python train.py --ds ETTh1 --model autoformer

References

• Informer
• Autoformer

Citation

@inproceedings{haoyietal-informer-2021,
  author    = {Haoyi Zhou and
               Shanghang Zhang and
               Jieqi Peng and
               Shuai Zhang and
               Jianxin Li and
               Hui Xiong and
               Wancai Zhang},
  title     = {Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting},
  booktitle = {The Thirty-Fifth {AAAI} Conference on Artificial Intelligence, {AAAI} 2021, Virtual Conference},
  volume    = {35},
  number    = {12},
  pages     = {11106--11115},
  publisher = {{AAAI} Press},
  year      = {2021},
}

@inproceedings{wu2021autoformer,
  title={Autoformer: Decomposition Transformers with {Auto-Correlation} for Long-Term Series Forecasting},
  author={Haixu Wu and Jiehui Xu and Jianmin Wang and Mingsheng Long},
  booktitle={Advances in Neural Information Processing Systems},
  year={2021}
}