metadata
dataset_info:
features:
- name: id
dtype: string
- name: imgs
dtype: image
- name: words
sequence: string
- name: bounding_boxes
sequence:
sequence: int64
- name: query
dtype: string
- name: answers
dtype: string
splits:
- name: train
num_bytes: 120380268
num_examples: 135
- name: test
num_bytes: 13362484
num_examples: 15
download_size: 133288363
dataset_size: 133742752
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*