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
- name: category
dtype: string
- name: answer_text
dtype: string
splits:
- name: train
num_bytes: 49547053
num_examples: 2000
download_size: 49100512
dataset_size: 49547053
configs:
- config_name: default
data_files:
- split: train
path: data/train-*