metadata
dataset_info:
features:
- name: comment_text
dtype: string
- name: toxic
dtype:
class_label:
names:
'0': 'false'
'1': 'true'
- name: severe_toxic
dtype:
class_label:
names:
'0': 'false'
'1': 'true'
- name: obscene
dtype:
class_label:
names:
'0': 'false'
'1': 'true'
- name: threat
dtype:
class_label:
names:
'0': 'false'
'1': 'true'
- name: insult
dtype:
class_label:
names:
'0': 'false'
'1': 'true'
- name: identity_hate
dtype:
class_label:
names:
'0': 'false'
'1': 'true'
splits:
- name: train
num_bytes: 71282294
num_examples: 159571
- name: test
num_bytes: 28241963
num_examples: 63978
download_size: 57648132
dataset_size: 99524257
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*