metadata
dataset_info:
features:
- name: id
dtype: string
- name: comment_text
dtype: string
- name: toxic
dtype: int64
- name: severe_toxic
dtype: int64
- name: obscene
dtype: int64
- name: threat
dtype: int64
- name: insult
dtype: int64
- name: identity_hate
dtype: int64
- name: label
dtype: int64
splits:
- name: train
num_bytes: 95255681
num_examples: 201194
- name: valid
num_bytes: 10527948
num_examples: 22355
download_size: 61670038
dataset_size: 105783629
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*