PubmedQA dataset derived from the files contained or linked in the official pubmedqa repo:
https://github.com/pubmedqa/pubmedqa(modified from
https://huggingface.co/datasets/bigbio/pubmed_qa)
PubMedQA is a biomedical question answering (QA) dataset collected from PubMed abstracts. The task of PubMedQA is to answer research biomedical questions with yes/no/maybe using the corresponding abstracts. PubMedQA has 1k expert-annotated (PQA-L), 61.2k… See the full description on the dataset page:
https://huggingface.co/datasets/rschf/pubmedqa.
