Buckets:
457 MB
7 files
Updated about 2 months ago
Ctrl+K
| Name | Size | Uploaded | Xet hash |
|---|---|---|---|
| checkpoints | 2 items | ||
| references | 1 items | ||
| .gitattributes | 1.52 kB xet | 818ba6de | |
| DATASET.md | 1.74 kB xet | 95d569ea | |
| LICENSE | 6.49 kB xet | 499d5291 | |
| README.md | 14.6 kB xet | 685564a2 |
The Open Catalysts 2025 (OC25) Dataset, Evaluations, and Models
Dataset
LICENSE: The OC25 dataset is provided under a CC-BY-4.0 license
Download links and instructions on how to access the data can be found here.
Model checkpoints
LICENSE: OC25 model checkpoints are provided under the FAIR Chemistry License
| Name | Split | Model reference name | Checkpoint |
| eSEN-sm-conserving | All | esen-sm-conserving-all-oc25 | esen_sm_conserve.pt |
| eSEN-md-direct | All | esen-md-direct-all-oc25 | esen_md_direct.pt |
How to use
Model checkpoints can be readily used in the fairchem repo using our custom ASE-calculator.
You can find details about using the model, along with other utilities including tutorials and documentation in the official fairchem repo. Make sure you have a Hugging Face account, and have applied for model access to this repository.
You can use the following to save an authentication token:
huggingface-cli login
The model reference names in the table above can be used to set up the ASE-calculator. For example:
from fairchem.core import pretrained_mlip, FAIRChemCalculator
predictor = pretrained_mlip.get_predict_unit("esen-sm-conserving-all-oc25", device="cuda")
calc = FAIRChemCalculator(predictor)
Support
If you run into any issues feel free to post your questions or comments on Github Issues.
Citation
If you use this work, please cite:
@misc{oc25,
title={The Open Catalyst 2025 (OC25) Dataset and Models for Solid-Liquid Interfaces},
author={Sushree Jagriti Sahoo and Mikael Maraschin and Daniel S. Levine and Zachary Ulissi and C. Lawrence Zitnick and Joel B Varley and Joseph A. Gauthier and Nitish Govindarajan and Muhammed Shuaibi},
year={2025},
eprint={2509.17862},
archivePrefix={arXiv},
primaryClass={cond-mat.mtrl-sci},
url={https://arxiv.org/abs/2509.17862},
}
- Total size
- 457 MB
- Files
- 7
- Last updated
- Jun 12
- Pre-warmed CDN
- US EU US EU