---
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-*
---