metadata
dataset_info:
features:
- name: image
dtype: image
- name: solution
struct:
- name: text
dtype: string
- name: original_question
dtype: string
- name: original_answer
dtype: string
- name: question
dtype: string
- name: language
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 13757645092.564
num_examples: 5686
download_size: 13742148577
dataset_size: 13757645092.564
configs:
- config_name: default
data_files:
- split: train
path: data/train-*