metadata
dataset_info:
features:
- name: image
dtype: image
- name: example_id
dtype: string
- name: text
dtype: string
- name: split
dtype: string
splits:
- name: train
num_bytes: 4955429308
num_examples: 9647
- name: validation
num_bytes: 47944681
num_examples: 100
- name: test
num_bytes: 2618799457
num_examples: 5100
download_size: 7635877062
dataset_size: 7622173446
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*