metadata
dataset_info:
features:
- name: image_path
dtype: string
- name: caption
dtype: string
splits:
- name: train
num_bytes: 182738
num_examples: 1200
- name: test
num_bytes: 8112
num_examples: 50
download_size: 70630
dataset_size: 190850
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*