metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 1563360143
num_examples: 2000
- name: test
num_bytes: 571628301
num_examples: 799
download_size: 2134048674
dataset_size: 2134988444
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*