metadata
dataset_info:
features:
- name: image
dtype: image
- name: text
dtype: string
- name: prompt
dtype: string
splits:
- name: train
num_bytes: 2257467472
num_examples: 27000
- name: validation
num_bytes: 122247646
num_examples: 1500
- name: test
num_bytes: 122151760
num_examples: 1500
download_size: 2478547830
dataset_size: 2501866878
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*