metadata
dataset_info:
features:
- name: image
dtype: image
- name: prompt
dtype: string
- name: json_output
dtype: string
splits:
- name: train
num_bytes: 736367161
num_examples: 2684
- name: validation
num_bytes: 158028059
num_examples: 576
- name: test
num_bytes: 158028181
num_examples: 576
download_size: 1039279845
dataset_size: 1052423401
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*