nfflr.data.mlearn_dataset#
- nfflr.data.mlearn_dataset(elements: str | list[str] = 'Si', transform: Module | None = None, diskcache: bool = False)[source]#
Construct mlearn dataset with standard splits.
Downloads and caches json datafiles from github.com/materialsvirtuallab/mlearn to nfflr.CACHE directory, which respects XDG_CACHE_HOME.