SumTablets

SumTablets 🏺 A Transliteration Dataset of Sumerian Tablets Preprocessing scripts on GitHub. We welcome contributions! What is it? SumTablets is a dataset of 91,606 Sumerian cuneiform tablets, structured as glyph–transliteration pairs. It is designed for the task of Sumerian transliteration, a conventional system for rendering an interpretation of a tablet using the Latin alphabet. We build upon existing data with the aim of providing: Structure. The transliterations… See the full description on the dataset page:

Source: Hugging Face Hub (colesimmons/SumTablets). Metadata imported from the dataset’s Hub tags.