metadata
dataset_info:
features:
- name: text_id
dtype: int64
- name: caption
dtype: string
splits:
- name: train
num_bytes: 36551702
num_examples: 566747
- name: val
num_bytes: 1610843
num_examples: 25010
- name: test
num_bytes: 1610345
num_examples: 25010
download_size: 21814166
dataset_size: 39772890