metadata
dataset_info:
features:
- name: source
dtype: string
splits:
- name: train
num_bytes: 3295598692
num_examples: 746337
- name: validation
num_bytes: 358484072
num_examples: 82903
download_size: 2108912451
dataset_size: 3654082764
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*