metadata
dataset_info:
features:
- name: image
dtype: image
- name: image_id
dtype: string
- name: caption
dtype: string
- name: cui
sequence: string
- name: questions
sequence: string
splits:
- name: test
num_bytes: 2588748056.49
num_examples: 9927
download_size: 2585818733
dataset_size: 2588748056.49
configs:
- config_name: default
data_files:
- split: test
path: data/test-*