metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: large_string
splits:
- name: train
num_bytes: 696637687
num_examples: 54089
- name: test
num_bytes: 13264389
num_examples: 1025
download_size: 681172216
dataset_size: 709902076
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*