metadata
dataset_info:
features:
- name: id
dtype: int64
- name: file_path
dtype: string
- name: captions
sequence: string
- name: split
dtype: string
- name: image
dtype: image
- name: height
dtype: int64
- name: width
dtype: int64
splits:
- name: train
num_bytes: 376349058.71
num_examples: 18505
- name: test
num_bytes: 31532243
num_examples: 1000
- name: val
num_bytes: 21807205
num_examples: 1000
download_size: 506086048
dataset_size: 429688506.71
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: val
path: data/val-*