metadata
dataset_info:
features:
- name: words
sequence: string
- name: norm_bboxes
sequence:
sequence: float64
- name: ner_tags
sequence: 'null'
- name: class
dtype:
class_label:
names:
'0': toc
'1': entry
'2': other
- name: image
dtype: image
splits:
- name: train
num_bytes: 166520938.02956522
num_examples: 862
- name: test
num_bytes: 57215447.970434785
num_examples: 288
download_size: 209131993
dataset_size: 223736386