metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: format
dtype: string
splits:
- name: train
num_bytes: 2763352331.275
num_examples: 50801
- name: validation
num_bytes: 322237958.496
num_examples: 5976
- name: test
num_bytes: 161977118.128
num_examples: 2988
download_size: 3228895917
dataset_size: 3247567407.8989997
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*