metadata
dataset_info:
features:
- name: id
dtype: int32
- name: image
dtype: image
- name: choices
list: string
- name: answer_type
dtype: string
- name: answers
struct:
- name: A
dtype: float32
- name: B
dtype: float32
- name: C
dtype: float32
- name: D
dtype: float32
- name: E
dtype: float32
- name: year
dtype: int32
splits:
- name: train
num_bytes: 212298417
num_examples: 400
download_size: 211832329
dataset_size: 212298417
configs:
- config_name: default
data_files:
- split: train
path: data/train-*