kyegomez/Aurora

Implementation of the paper: "Aurora: A Foundation Model of the Atmosphere" in PyTorch

39
/ 100
Emerging

Built on a Swin Transformer UNet 3D architecture, Aurora processes spatiotemporal atmospheric data through hierarchical shifted-window attention mechanisms optimized for weather prediction tasks. The model operates on 5D tensor inputs (batch, depth, height, width, channels) representing volumetric atmospheric fields, enabling efficient multi-scale feature extraction across spatial and temporal dimensions. Installable via pip as `aurora-torch`, it integrates directly into PyTorch workflows for atmospheric modeling and weather forecasting applications.

No Package No Dependents
Maintenance 13 / 25
Adoption 6 / 25
Maturity 9 / 25
Community 11 / 25

How are scores calculated?

Stars

22

Forks

3

Language

Python

License

MIT

Last pushed

Mar 09, 2026

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/agents/kyegomez/Aurora"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.