metadata
dataset_info:
features:
- name: ID
dtype: string
- name: image_path
dtype: string
- name: topic
dtype: string
- name: prompt
dtype: string
- name: ground_truth
dtype: string
- name: expected_bias
dtype: string
- name: type_of_question
dtype: string
- name: pixel
dtype: int64
- name: image
dtype: image
splits:
- name: train
num_bytes: 422190579
num_examples: 546
download_size: 216930631
dataset_size: 422190579
configs:
- config_name: default
data_files:
- split: train
path: data/train-*