metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: text
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 5780469153.75
num_examples: 26002
download_size: 5774050664
dataset_size: 5780469153.75
configs:
- config_name: default
data_files:
- split: train
path: data/train-*