---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: text
    dtype: string
  - name: url
    dtype: string
  - name: perplexity_kenlm
    dtype: int64
  - name: label_identity_attack
    dtype: float64
  - name: label_insult
    dtype: float64
  - name: label_obscene
    dtype: float64
  - name: label_severe_toxicity
    dtype: float64
  - name: label_threat
    dtype: float64
  - name: label_toxicity
    dtype: float64
  splits:
  - name: train
    num_bytes: 58313946485
    num_examples: 5110923
  download_size: 30645534064
  dataset_size: 58313946485
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---