metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: image
dtype: image
- name: caption
dtype: string
splits:
- name: train
num_bytes: 267160698
num_examples: 316
download_size: 267063112
dataset_size: 267160698
configs:
- config_name: default
data_files:
- split: train
path: data/train-*