metadata
dataset_info:
features:
- name: image
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 60793590.988
num_examples: 2196
- name: test
num_bytes: 16076318
num_examples: 546
- name: validation
num_bytes: 15781217
num_examples: 546
download_size: 30868515
dataset_size: 92651125.988
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*