metadata
dataset_info:
features:
- name: image
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 407706056
num_examples: 660
- name: test
num_bytes: 177834801
num_examples: 284
download_size: 467315076
dataset_size: 585540857
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*