metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: text
dtype: string
- name: image
dtype: binary
splits:
- name: train
num_bytes: 1678971
num_examples: 2
download_size: 1680917
dataset_size: 1678971
configs:
- config_name: default
data_files:
- split: train
path: data/train-*