Datasets:

Hieuman
/

Toxic-conversation-reduced

dataset_info:
  features:
    - name: query
      dtype: string
    - name: positive
      sequence: string
    - name: negative
      sequence: string
    - name: cluster
      dtype: int64
  splits:
    - name: train
      num_bytes: 6963004.556419191
      num_examples: 20000
  download_size: 3812022
  dataset_size: 6963004.556419191
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*