metadata
dataset_info:
features:
- name: image
dtype: image
- name: id
dtype: string
- name: query
struct:
- name: question
dtype: string
- name: answers
sequence: string
- name: words
sequence: string
- name: bounding_boxes
sequence:
sequence: int64
splits:
- name: train
num_bytes: 1046177127.1854271
num_examples: 1100
- name: test
num_bytes: 467926343.91457283
num_examples: 492
download_size: 1510652483
dataset_size: 1514103471.1
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*