metadata
dataset_info:
features:
- name: image
dtype: image
- name: file_name
dtype: string
- name: query
dtype: string
- name: answer
dtype: string
splits:
- name: train
num_bytes: 5792004
num_examples: 70
download_size: 1155766
dataset_size: 5792004
configs:
- config_name: default
data_files:
- split: train
path: data/train-*