metadata
dataset_info:
features:
- name: id
dtype: string
- name: question_id
dtype: string
- name: question
dtype: string
- name: answer
dtype: string
- name: image_source
dtype: string
- name: image
dtype: image
- name: category
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 74355758
num_examples: 467
download_size: 67654348
dataset_size: 74355758
configs:
- config_name: default
data_files:
- split: train
path: data/train-*