metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 23095195
num_examples: 2000
- name: test
num_bytes: 61214498
num_examples: 3000
download_size: 68105850
dataset_size: 84309693
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*