metadata
dataset_info:
features:
- name: id
dtype: uint32
- name: toxic_sentence
dtype: string
- name: neutral_sentence
dtype: string
splits:
- name: train
num_bytes: 2011316
num_examples: 16828
download_size: 1451861
dataset_size: 2011316
configs:
- config_name: default
data_files:
- split: train
path: data/train-*