metadata
dataset_info:
features:
- name: image
dtype: image
- name: captions
sequence: string
- name: class
dtype: string
- name: family
dtype: string
splits:
- name: train
num_bytes: 193847123.16
num_examples: 1680
- name: validation
num_bytes: 47847281
num_examples: 420
download_size: 240885008
dataset_size: 241694404.16
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*