This is a novel multiple choice, clinical question & answering (Q&A) dataset in Swedish consisting of
3,180 questions. The dataset was created from a series of exams aimed at evaluating doctors’ clinical understanding
and decision making and is the first open-source clinical Q&A dataset in Swedish. The exams – originally in PDF
format – were parsed and each question manually checked and curated in order to limit errors in the dataset.
Please read the… See the full description on the dataset page:
https://huggingface.co/datasets/nicher92/medqa-swe.