metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: writer
dtype: string
- name: form
dtype: string
- name: paragraph
dtype: int64
splits:
- name: train
num_bytes: 97658800
num_examples: 4821
- name: test
num_bytes: 19042122
num_examples: 955
download_size: 116234496
dataset_size: 116700922
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*