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: 34057432
num_examples: 2101
- name: validation
num_bytes: 4673256
num_examples: 300
- name: test
num_bytes: 9645474
num_examples: 601
download_size: 47964261
dataset_size: 48376162
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*