metadata
dataset_info:
features:
- name: image
dtype: image
- name: question
dtype: string
- name: answer
dtype: string
- name: explanation
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 42370995
num_examples: 1000
- name: validation
num_bytes: 3907602
num_examples: 100
download_size: 28767519
dataset_size: 46278597
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*