metadata
dataset_info:
features:
- name: id
dtype: string
- name: media
list: image
- name: messages
dtype: string
- name: answer
dtype: string
- name: question_type
dtype: string
splits:
- name: train
num_bytes: 393384208
num_examples: 4919
- name: validation
num_bytes: 81826982
num_examples: 1053
- name: test
num_bytes: 90948107
num_examples: 1061
download_size: 565126493
dataset_size: 566159297
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*