metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
splits:
- name: train
num_bytes: 63623143
num_examples: 2000
- name: validation
num_bytes: 1206261
num_examples: 40
- name: test
num_bytes: 6097271
num_examples: 200
download_size: 60857730
dataset_size: 70926675
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*