metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: generated_caption
dtype: string
splits:
- name: train
num_bytes: 2467169489.125
num_examples: 7999
download_size: 2418777187
dataset_size: 2467169489.125