metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 59921589
num_examples: 2100
- name: val
num_bytes: 25922766.5
num_examples: 900
download_size: 84801147
dataset_size: 85844355.5