metadata
dataset_info:
features:
- name: text
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 3735636754
num_examples: 12000
download_size: 3733205906
dataset_size: 3735636754
configs:
- config_name: default
data_files:
- split: train
path: data/train-*