metadata
dataset_info:
features:
- name: image_id
dtype: string
- name: images
dtype: image
- name: problem
dtype: string
- name: question_only
dtype: string
- name: answer
dtype: string
- name: answer_only
dtype: string
- name: category
dtype: string
- name: level
dtype: string
- name: question_with_options
dtype: string
- name: options
dtype: string
- name: rating
dtype: int32
splits:
- name: train
num_bytes: 244820495.5
num_examples: 5396
- name: val
num_bytes: 22771643
num_examples: 531
download_size: 253429496
dataset_size: 267592138.5
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*