README.md

metadata

dataset_info:
  features:
    - name: anchor
      dtype: string
    - name: positive
      dtype: string
    - name: negative
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 1281750763
      num_examples: 3401694
  download_size: 811108906
  dataset_size: 1281750763
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*