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'
- name: prediction_hate
dtype: float64
- name: prediction_nsfw
dtype: float64
splits:
- name: test
num_bytes: 29265611
num_examples: 63978
download_size: 17145033
dataset_size: 29265611
configs:
- config_name: default
data_files:
- split: test
path: data/test-*