metadata
dataset_info:
features:
- name: target
dtype: image
- name: source
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 202903479
num_examples: 300
download_size: 201481506
dataset_size: 202903479
configs:
- config_name: default
data_files:
- split: train
path: data/train-*