wiki_bio

  • Description:

WikiBio is constructed using Wikipedia biography pages, it contains the first paragraph and the infobox tokenized. The dataset follows a standarized table format.

Split Examples
'test' 72,831
'train' 582,659
'validation' 72,831
  • Feature structure:
FeaturesDict({
    'input_text': FeaturesDict({
        'context': string,
        'table': Sequence({
            'column_header': string,
            'content': string,
            'row_number': int16,
        }),
    }),
    'target_text': string,
})
  • Feature documentation:
Feature Class Shape Dtype Description
FeaturesDict
input_text FeaturesDict
input_text/context Tensor string
input_text/table Sequence
input_text/table/column_header Tensor string
input_text/table/content Tensor string
input_text/table/row_number Tensor int16
target_text Tensor string
  • Citation:
@inproceedings{lebret-etal-2016-neural,
    title = "Neural Text Generation from Structured Data with Application to the Biography Domain",
    author = "Lebret, R{'e}mi  and
      Grangier, David  and
      Auli, Michael",
    booktitle = "Proceedings of the 2016 Conference on Empirical Methods in Natural Language Processing",
    month = nov,
    year = "2016",
    address = "Austin, Texas",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/D16-1128",
    doi = "10.18653/v1/D16-1128",
    pages = "1203--1213",
}