metadata
dataset_info:
features:
- name: image
dtype: image
- name: captions
sequence: string
splits:
- name: train
num_bytes: 406958641.136
num_examples: 1472
- name: validation
num_bytes: 99512628
num_examples: 391
- name: test
num_bytes: 107370462
num_examples: 487
download_size: 548327841
dataset_size: 613841731.1359999
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*