metadata
dataset_info:
features:
- name: image
dtype: image
- name: image_id
dtype: string
- name: caption
dtype: string
splits:
- name: train
num_bytes: 12809452439.625
num_examples: 65419
- name: validation
num_bytes: 277877322.125
num_examples: 8175
- name: test
num_bytes: 275221393
num_examples: 8176
download_size: 13356690345
dataset_size: 13362551154.75
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*