Awesome

Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting (AAAI'21 Best Paper)

This is the origin Pytorch implementation of Informer in the following paper: Informer: Beyond Efficient Transformer for Long Sequence Time-Series Forecasting. Special thanks to Jieqi Peng@cookieminions for building this repo.

:triangular_flag_on_post:News(Mar 27, 2023): We will release Informer V2 soon.

:triangular_flag_on_post:News(Feb 28, 2023): The Informer's extension paper is online on AIJ.

:triangular_flag_on_post:News(Mar 25, 2021): We update all experiment results with hyperparameter settings.

:triangular_flag_on_post:News(Feb 22, 2021): We provide Colab Examples for friendly usage.

:triangular_flag_on_post:News(Feb 8, 2021): Our Informer paper has been awarded AAAI'21 Best Paper [Official][Beihang][Rutgers]! We will continue this line of research and update on this repo. Please star this repo and cite our paper if you find our work is helpful for you.

ProbSparse Attention

The self-attention scores form a long-tail distribution, where the "active" queries lie in the "head" scores and "lazy" queries lie in the "tail" area. We designed the ProbSparse Attention to select the "active" queries rather than the "lazy" queries. The ProbSparse Attention with Top-u queries forms a sparse Transformer by the probability distribution. Why not use Top-u keys? The self-attention layer's output is the re-represent of input. It is formulated as a weighted combination of values w.r.t. the score of dot-product pairs. The top queries with full keys encourage a complete re-represent of leading components in the input, and it is equivalent to selecting the "head" scores among all the dot-product pairs. If we choose Top-u keys, the full keys just preserve the trivial sum of values within the "long tail" scores but wreck the leading components' re-represent.

Requirements

• Python 3.6
• matplotlib == 3.1.1
• numpy == 1.19.4
• pandas == 0.25.1
• scikit_learn == 0.21.3
• torch == 1.8.0

Dependencies can be installed using the following command:

pip install -r requirements.txt

Data

The ETT dataset used in the paper can be downloaded in the repo ETDataset. The required data files should be put into data/ETT/ folder. A demo slice of the ETT data is illustrated in the following figure. Note that the input of each dataset is zero-mean normalized in this implementation.

The ECL data and Weather data can be downloaded here.

• Google Drive
• BaiduPan, password: 6gan

Reproducibility

To easily reproduce the results you can follow the next steps:

1. Initialize the docker image using: make init.
2. Download the datasets using: make dataset.
3. Run each script in scripts/ using make run_module module="bash ETTh1.sh" for each script.
4. Alternatively, run all the scripts at once:

for file in `ls scripts`; do make run_module module="bash scripts/$script"; done

Usage

<span id="colablink">Colab Examples:</span> We provide google colabs to help reproduce and customize our repo, which includes experiments(train and test), prediction, visualization and custom data.

Commands for training and testing the model with ProbSparse self-attention on Dataset ETTh1, ETTh2 and ETTm1 respectively:

# ETTh1
python -u main_informer.py --model informer --data ETTh1 --attn prob --freq h

# ETTh2
python -u main_informer.py --model informer --data ETTh2 --attn prob --freq h

# ETTm1
python -u main_informer.py --model informer --data ETTm1 --attn prob --freq t

More parameter information please refer to main_informer.py.

We provide a more detailed and complete command description for training and testing the model:

python -u main_informer.py --model <model> --data <data>
--root_path <root_path> --data_path <data_path> --features <features>
--target <target> --freq <freq> --checkpoints <checkpoints>
--seq_len <seq_len> --label_len <label_len> --pred_len <pred_len>
--enc_in <enc_in> --dec_in <dec_in> --c_out <c_out> --d_model <d_model>
--n_heads <n_heads> --e_layers <e_layers> --d_layers <d_layers>
--s_layers <s_layers> --d_ff <d_ff> --factor <factor> --padding <padding>
--distil --dropout <dropout> --attn <attn> --embed <embed> --activation <activation>
--output_attention --do_predict --mix --cols <cols> --itr <itr>
--num_workers <num_workers> --train_epochs <train_epochs>
--batch_size <batch_size> --patience <patience> --des <des>
--learning_rate <learning_rate> --loss <loss> --lradj <lradj>
--use_amp --inverse --use_gpu <use_gpu> --gpu <gpu> --use_multi_gpu --devices <devices>

The detailed descriptions about the arguments are as following:

<span id="resultslink">Results</span>

We have updated the experiment results of all methods due to the change in data scaling. We are lucky that Informer gets performance improvement. Thank you @lk1983823 for reminding the data scaling in issue 41.

Besides, the experiment parameters of each data set are formated in the .sh files in the directory ./scripts/. You can refer to these parameters for experiments, and you can also adjust the parameters to obtain better mse and mae results or draw better prediction figures.

FAQ

If you run into a problem like RuntimeError: The size of tensor a (98) must match the size of tensor b (96) at non-singleton dimension 1, you can check torch version or modify code about Conv1d of TokenEmbedding in models/embed.py as the way of circular padding mode in Conv1d changed in different torch versions.

<span id="citelink">Citation</span>

If you find this repository useful in your research, please consider citing the following papers:

@article{haoyietal-informerEx-2023,
  author    = {Haoyi Zhou and
               Jianxin Li and
               Shanghang Zhang and
               Shuai Zhang and
               Mengyi Yan and
               Hui Xiong},
  title     = {Expanding the prediction capacity in long sequence time-series forecasting},
  journal   = {Artificial Intelligence},
  volume    = {318},
  pages     = {103886},
  issn      = {0004-3702},
  year      = {2023},
}

@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},
}

Contact

If you have any questions, feel free to contact Haoyi Zhou through Email (zhouhaoyi1991@gmail.com) or Github issues. Pull requests are highly welcomed!

Acknowledgments

Thanks for the computing infrastructure provided by Beijing Advanced Innovation Center for Big Data and Brain Computing (BDBC). At the same time, thank you all for your attention to this work!