metadata
dataset_info:
features:
- name: images
dtype: image
- name: correct_caption
dtype: string
- name: hard_text_1
dtype: string
- name: hard_text_2
dtype: string
- name: hard_text_3
dtype: string
- name: hard_text_4
dtype: string
- name: true_triples
list:
list: string
- name: false_triples_1
list:
list: string
- name: false_triples_2
list:
list: string
- name: false_triples_3
list:
list: string
- name: false_triples_4
list:
list: string
splits:
- name: train
num_bytes: 1728875080
num_examples: 5000
download_size: 1667784936
dataset_size: 1728875080
configs:
- config_name: default
data_files:
- split: train
path: data/train-*