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: 2748165580
num_examples: 19654
- name: validation
num_bytes: 871504371
num_examples: 6259
- name: test
num_bytes: 934330711
num_examples: 6719
download_size: 4549836684
dataset_size: 4554000662
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*