metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 23026710206.8
num_examples: 142291
- name: validation
num_bytes: 1459190963.464
num_examples: 9036
download_size: 21250472921
dataset_size: 24485901170.264
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*