An Efficient finetuning approach for Extended GraphCast.
This project aims to provide an open-source and user-friendly framework to use and enhance GraphCast, a powerful ML weather forecasting model. Our focus is on simplifying the finetuning process and significantly expanding the output to include a broader range of weather variables.
We would like to thank:
- The authors of the paper GraphCast: Learning skillful medium-range global weather forecasting from which we are basing this project.
For now, this project is conducted by: Lounès Meddahi.