metadata
dataset_info:
features:
- name: caption
dtype: string
- name: image
dtype: image
splits:
- name: train
num_bytes: 315307268
num_examples: 999
download_size: 288501432
dataset_size: 315307268
configs:
- config_name: default
data_files:
- split: train
path: data/train-*