metadata
dataset_info:
features:
- name: words
sequence: string
- name: labels
sequence: int64
- name: boxes
sequence:
sequence: int64
- name: file_name
dtype: string
- name: actual_bboxes
sequence:
sequence: int64
- name: image
sequence:
sequence:
sequence: uint8
splits:
- name: train
num_bytes: 9991089032
num_examples: 65151
- name: val
num_bytes: 1110120734
num_examples: 7239
- name: test
num_bytes: 6085345877
num_examples: 39682
download_size: 7559946719
dataset_size: 17186555643