metadata
dataset_info:
features:
- name: text
dtype: string
- name: id
dtype: string
- name: source
dtype: string
- name: safety_score
dtype: int8
- name: has_annotation
dtype: bool
- name: is_bad
dtype: bool
splits:
- name: train
num_bytes: 841481589
num_examples: 100000
download_size: 487668402
dataset_size: 841481589
configs:
- config_name: default
data_files:
- split: train
path: data/train-*