metadata
dataset_info:
features:
- name: docid
dtype: string
- name: title
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 13936141152
num_examples: 32893221
download_size: 8002214716
dataset_size: 13936141152
configs:
- config_name: default
data_files:
- split: train
path: data/train-*