WeatherNext Graph and Gen are available as open source packages on GitHub.
The package contains example code to run and train the weather models used in the research papers GraphCast and GenCast. It also provides pretrained model weights, normalization statistics and example input data on Google Cloud Bucket.
Full model training requires downloading the ERA5 dataset, available from ECMWF. This can best be accessed as Zarr from Weatherbench2's ERA5 data. Data for operational fine-tuning can similarly be accessed at Weatherbench2's HRES 0th frame data.
Where to start
Multiple models are available including operational and high/low resolution models.
The best starting point is to open the demo notebooks in Colaboratory, which gives an example of loading data, generating random weights or loading a pre-trained snapshot, generating predictions, computing the loss and computing gradients.
See GitHub for more details on available models and relevant files.
License
License information for the open source models are detailed on GitHub
Citations
If you use this work, consider citing our papers:
@article{lam2023learning,
title={Learning skillful medium-range global weather forecasting},
author={Lam, Remi and Sanchez-Gonzalez, Alvaro and Willson, Matthew and Wirnsberger, Peter and Fortunato, Meire and Alet, Ferran and Ravuri, Suman and Ewalds, Timo and Eaton-Rosen, Zach and Hu, Weihua and others},
journal={Science},
volume={382},
number={6677},
pages={1416--1421},
year={2023},
publisher={American Association for the Advancement of Science}
}
@article{Price2025,
author = {Price, Ilan and Alet, Ferran and El-Kadi, Andrew and Masters, Dominic and Markou, Stratis and Andersson, Tom R. and Stott, Jacklynn and Lam, Remi and Willson, Matthew and Sanchez-Gonzalez, Alvaro and Battaglia, Peter},
title = {Probabilistic weather forecasting with machine learning},
journal = {Nature},
year = {2025},
doi = {10.1038/s41586-024-08252-9},
url = {https://doi.org/10.1038/s41586-024-08252-9}
}
@misc{alet2025skillful,
title={Skillful joint probabilistic weather forecasting from marginals},
author={Ferran Alet and Ilan Price and Andrew El-Kadi and Dominic Masters and Stratis Markou and Tom R. Andersson and Jacklynn Stott and Remi Lam and Matthew Willson and Alvaro Sanchez-Gonzalez and Peter Battaglia},
year={2025},
eprint={2506.10772},
archivePrefix={arXiv},
primaryClass={cs.LG},
url={https://arxiv.org/abs/2506.10772}
}