README.md

metadata

configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: validation
        path: data/validation-*
dataset_info:
  features:
    - name: sentence1
      dtype: string
    - name: sentence2
      dtype: string
    - name: label
      dtype: string
  splits:
    - name: train
      num_bytes: 337045633
      num_examples: 1565996
    - name: test
      num_bytes: 18725973
      num_examples: 87000
    - name: validation
      num_bytes: 18760874
      num_examples: 87000
  download_size: 261610286
  dataset_size: 374532480

Dataset Card for "discoveryhard"

https://github.com/sileod/Discovery

@inproceedings{sileo-etal-2019-mining,
    title = "Mining Discourse Markers for Unsupervised Sentence Representation Learning",
    author = "Sileo, Damien  and
      Van De Cruys, Tim  and
      Pradel, Camille  and
      Muller, Philippe",
    booktitle = "Proceedings of the 2019 Conference of the North {A}merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers)",
    month = jun,
    year = "2019",
    address = "Minneapolis, Minnesota",
    publisher = "Association for Computational Linguistics",
    url = "https://www.aclweb.org/anthology/N19-1351",
    pages = "3477--3486",
}