metadata
dataset_info:
features:
- name: image
dtype: image
- name: question_type
dtype: string
- name: answers
sequence: string
- name: id_image
dtype: int64
- name: question_id
dtype: int64
- name: question
dtype: string
splits:
- name: test
num_bytes: 833577670
num_examples: 5046
- name: train
num_bytes: 1487594256
num_examples: 9009
download_size: 2315635688
dataset_size: 2321171926
configs:
- config_name: default
data_files:
- split: test
path: data/test-*
- split: train
path: data/train-*