metadata
dataset_info:
features:
- name: images
dtype: image
- name: image_description
dtype: string
splits:
- name: train
num_bytes: 2824689490
num_examples: 2400
download_size: 2824222854
dataset_size: 2824689490
configs:
- config_name: default
data_files:
- split: train
path: data/train-*