metadata
dataset_info:
features:
- name: image
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 182181560.5003668
num_examples: 3315
- name: test
num_bytes: 44699653.954633206
num_examples: 829
download_size: 212651682
dataset_size: 226881214.455
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*