metadata
dataset_info:
features:
- name: title
dtype: string
- name: article
dtype: string
- name: label
dtype: string
- name: prompt
dtype: string
- name: text_token_length
dtype: int64
- name: seed_data
dtype: string
- name: format
dtype: string
- name: audience
dtype: string
splits:
- name: train
num_bytes: 4337125495
num_examples: 1067358
download_size: 2828980035
dataset_size: 4337125495
configs:
- config_name: default
data_files:
- split: train
path: data/train-*