metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 60392465.7
num_examples: 2100
- name: val
num_bytes: 25916253.5
num_examples: 900
download_size: 84975483
dataset_size: 86308719.2