metadata
dataset_info:
features:
- name: image_link
dtype: string
- name: group_id
dtype: int64
- name: entity_name
dtype: string
- name: answer
dtype: string
- name: question
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 81758
num_examples: 500
download_size: 15868
dataset_size: 81758
configs:
- config_name: default
data_files:
- split: train
path: data/train-*