metadata
dataset_info:
features:
- name: image
dtype: image
- name: image_id
dtype: string
- name: caption
dtype: string
- name: cui
sequence: 'null'
splits:
- name: train
num_bytes: 49800878.13333333
num_examples: 94
- name: validation
num_bytes: 10168445.666666666
num_examples: 20
- name: test
num_bytes: 11597571.199999997
num_examples: 21
download_size: 69531794
dataset_size: 71566895
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*