---
dataset_info:
  features:
  - name: title
    dtype: string
  - name: namespace
    dtype: string
  - name: timestamp
    dtype: string
  - name: text
    dtype: string
  - name: language
    dtype: string
  - name: toxicity
    dtype: bool
  - name: toxic-bert
    struct:
    - name: label
      dtype: string
    - name: score
      dtype: float64
  splits:
  - name: train
    num_bytes: 1979578
    num_examples: 3725
  download_size: 1017192
  dataset_size: 1979578
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---