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
splits:
- name: train
num_bytes: 6887148
num_examples: 1078
download_size: 2658905
dataset_size: 6887148
configs:
- config_name: default
data_files:
- split: train
path: data/train-*