MLCast is an open-source community building machine learning models for weather nowcasting. Researchers, developers, students, and weather enthusiasts contribute code, datasets, documentation, and compute—and everyone is welcome.
Why Collaborate¶
Build better models together. Many perspectives make models more accurate, reliable, and innovative.
Share resources and expertise. Institutions with GPUs train large models; others contribute software, data validation, evaluation, or docs. No specialized hardware required.
Accelerate innovation. Build on each other’s work instead of starting from scratch. Community feedback catches bugs and drives new features.
Learn and grow. Gain hands-on experience with ML, software development, weather science, and open-source collaboration—from your first pull request to advanced model architectures.
Open future for weather AI. Models, datasets, and software developed transparently and shared freely, so anyone can benefit.
Getting Started¶
Explore the MLCast GitHub organization and read the docs to see how the community is organized.
Follow the Get Started guide to set up the development environment.
Pick up a good first issue—designed to help new contributors learn the codebase while making meaningful contributions.
Discuss larger changes early on Slack to avoid duplicate work and align with project goals.
Create a branch, make and test your changes, and open a pull request. Maintainers review, give feedback, and guide it through.
With GPUs? Train and benchmark models. Without? Contribute through development, documentation, testing, and code reviews.
Connect¶
Join our Slack — questions, discussions, and real-time help.
Subscribe to the mailing list — send a blank email from the address you want to subscribe.
GitHub organization — repositories, issues, and pull requests.
Community & News — monthly community meetings and meeting notes.
Every contribution, no matter how small, moves the project forward.