metadata
dataset_info:
features:
- name: queries
sequence: string
- name: labels
sequence: string
- name: subquestions
list:
- name: answer
dtype: string
- name: question
dtype: string
- name: image
dtype: image
- name: base_question
dtype: string
- name: base_answer
dtype: string
splits:
- name: train
num_bytes: 69694686.625
num_examples: 1619
download_size: 68868615
dataset_size: 69694686.625
configs:
- config_name: default
data_files:
- split: train
path: data/train-*