metadata
dataset_info:
features:
- name: words
sequence: string
- name: ner_tags
sequence:
class_label:
names:
'0': other
'1': header
'2': answer
'3': question
- name: bboxes
sequence:
sequence: int64
- name: image
dtype: image
splits:
- name: test
num_bytes: 9944297
num_examples: 50
- name: train
num_bytes: 26607767
num_examples: 149
download_size: 140790319
dataset_size: 36552064
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*