metadata
dataset_info:
features:
- name: image
dtype: image
- name: image_path
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: id
dtype: string
splits:
- name: train
num_bytes: 57563969
num_examples: 370
download_size: 57515483
dataset_size: 57563969
configs:
- config_name: default
data_files:
- split: train
path: data/train-*