metadata
dataset_info:
features:
- name: document_id
dtype: string
- name: page_number
dtype: int32
- name: image
dtype: image
splits:
- name: train
num_bytes: 24585754452.049
num_examples: 64057
download_size: 24290141927
dataset_size: 24585754452.049
configs:
- config_name: default
data_files:
- split: train
path: data/train-*