A unified, verse-aligned multilingual Quran corpus spanning 79
languages and 185 translations. Every recension and translation is a
separate config (subset), all row-aligned on the canonical 6,236-ayah
verse_key (Hafs ʿan ʿAsim reading, 114 surahs).
The corpus also contains 111 tafsir configs: verse-grain classical
and openly licensed Arabic works, plus the native-passage and verse-expanded
views of Diyanet's Turkish Kur'an Yolu… See the full description on the dataset page:
https://huggingface.co/datasets/quranlab/quran.
