metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 90386900
num_examples: 2000
- name: test
num_bytes: 4472528
num_examples: 100
download_size: 31297145
dataset_size: 94859428
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*