Collection of Transformer based PyTorch Time Series models in the gluonts API.
pip3 install -U -r requirements.txtFor the code of the Lag-Llama paper, please use the lag-llama folder.
Note: The full version of the Lag-Llama paper is coming soon with comprehensive results and open-source model checkpoints. Stay Tuned!
For questions regarding the GitHub repository, please raise an issue here.
For questions regarding the Lag-Llama paper, please contact us at: kashif.rasul@gmail.com, arjun.ashok@servicenow.com.
If you use parts of this codebase, please cite the Lag-llama paper:
@misc{rasul2023lagllama,
title={Lag-Llama: Towards Foundation Models for Time Series Forecasting},
author={Kashif Rasul and Arjun Ashok and Andrew Robert Williams and Arian Khorasani and George Adamopoulos and Rishika Bhagwatkar and Marin Biloš and Hena Ghonia and Nadhir Vincent Hassen and Anderson Schneider and Sahil Garg and Alexandre Drouin and Nicolas Chapados and Yuriy Nevmyvaka and Irina Rish},
year={2023},
eprint={2310.08278},
archivePrefix={arXiv},
primaryClass={cs.LG}
}