A fully multilingual medical question-answering dataset covering 31 medical specialties with 35,481 clinical Q&A pairs, each provided in English, French, and Moroccan Darija. This is the v2 release, with every field filled — no more gaps. Part of the BRAIN HEALTH project, for research and educational use around multilingual clinical NLP.

What's new in v2

Compared to v1:

Improvement v1 v2

Columns 12 14 (added context_question_en… See the full description on the dataset page: https://huggingface.co/datasets/Williamsanderson/MedQAData-v2.