metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: images
list: image
- name: answer
dtype: string
splits:
- name: train
num_bytes: 81124028
num_examples: 50
download_size: 81126829
dataset_size: 81124028
configs:
- config_name: default
data_files:
- split: train
path: data/train-*