metadata
dataset_info:
features:
- name: image
dtype: image
- name: category
dtype: string
- name: question
dtype: string
- name: model_answer
dtype: string
- name: actual_answer
dtype: string
splits:
- name: train
num_bytes: 2867198
num_examples: 60
download_size: 920005
dataset_size: 2867198
configs:
- config_name: default
data_files:
- split: train
path: data/train-*