metadata
dataset_info:
features:
- name: image
dtype: image
- name: prompt
dtype: string
- name: json_output
dtype: string
splits:
- name: train
num_bytes: 689940985
num_examples: 2348
- name: validation
num_bytes: 148096473
num_examples: 504
- name: test
num_bytes: 148096407
num_examples: 504
download_size: 974361822
dataset_size: 986133865
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*