README.md

metadata

dataset_info:
  features:
    - name: context
      list: string
    - name: question
      dtype: string
    - name: answer
      dtype: string
  splits:
    - name: train
      num_bytes: 1330470319.419014
      num_examples: 4770
    - name: test
      num_bytes: 94276513.19992176
      num_examples: 338
  download_size: 748276863
  dataset_size: 1424746832.6189358
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*

Hard biomedical evaluation of question answering

Results of various models: https://huggingface.co/datasets/casperhansen/pmc-oa-markdown-qa-results