JARVIS-ML / CFID

JARVIS-ML / CFID is a research-grade descriptor benchmark dataset focused on classical force-field inspired descriptors and ML-ready materials property datasets for benchmarking materials informatics models. It is useful for graduate-level materials research, reproducible benchmarking, model training, validation, and comparison across computational, experimental, or characterization workflows.

Citation: JARVIS-ML / CFID. Materials dataset resource. https://jarvis.nist.gov/jarvisml

Acknowledgement: Curated by MaterialsAtlas Open Resources from public data portals, official project pages, benchmark suites, repositories, and publication-linked archives.