metadata
dataset_info:
features:
- name: img
dtype: image
- name: labels
sequence: string
- name: words
sequence: string
- name: bboxes
sequence:
sequence: int64
- name: filename
dtype: string
splits:
- name: train
num_bytes: 299073901
num_examples: 526
- name: val
num_bytes: 59447631
num_examples: 100
download_size: 319041399
dataset_size: 358521532