metadata
dataset_info:
features:
- name: id
dtype: string
- name: comment_text_bpe_tokenize_simple_dup_normalization
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: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 60937398
num_examples: 127656
- name: validation
num_bytes: 15066647
num_examples: 31915
- name: test
num_bytes: 29332039
num_examples: 63978
download_size: 59838960
dataset_size: 105336084
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*