README.md

metadata

dataset_info:
  features:
    - name: id
      dtype: uint32
    - name: toxic_sentence
      dtype: string
    - name: neutral_sentence
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 2264753
      num_examples: 16828
  download_size: 1398249
  dataset_size: 2264753
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*