metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: dataset_name
dtype: string
splits:
- name: train
num_bytes: 638330668911
num_examples: 4434226
download_size: 638243435255
dataset_size: 638330668911