CommonMorph's Datasets

Morphological Datasets

We are gathering an open-source, multilingual dataset of morphological information, freely available for training sub-word models. These datasets also play a vital role in documenting and preserving languages. The datasets are available in the Unimorph Format.