metadata
dataset_info:
features:
- name: file_path
dtype: string
- name: text
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 2581524871
num_examples: 23243
download_size: 2576957819
dataset_size: 2581524871
configs:
- config_name: default
data_files:
- split: train
path: data/train-*