OlmoEarth-v1_3

OlmoEarth-v1_3-Base is a ViT-Base (128M parameters) foundation model for remote sensing tasks over Sentinel-1, Sentinel-2, and Landsat images or image time series. It also includes a Perceiver (14.3M parameters) and a linear student which projects the 768 dimensional tokens down to 128 dimensions (0.1M parameters). This model (including the student) is used to generate OlmoEarth's global, annualized embeddings.

Please see our GitHub and paper for more information about OlmoEarth.

• GitHub: https://github.com/allenai/olmoearth_pretrain
• Paper: https://allenai.org/papers/olmoearth_v1_3

If you are interested in computing embeddings with OlmoEarth, or fine-tuning OlmoEarth for a downstream task, see our tutorials:

• Fine-tuning OlmoEarth for Segmentation using olmoearth_run
• Computing OlmoEarth Embeddings
• Fine-tuning OlmoEarth using rslearn

The OlmoEarth model is deployed in the OlmoEarth Platform, a solution that enables any organization to leverage our models and accelerate their mission, without requiring deep AI or engineering expertise. Sign up to get updates and partner access.

This model is licensed under the OlmoEarth Artifact License. It is intended for use in accordance with Ai2's Responsible Use Guidelines.