metadata
dataset_info:
features:
- name: id
dtype: int64
- name: prompt
dtype: string
- name: prompt_type
dtype: string
- name: qa_type
dtype: string
- name: answer_letter
dtype: string
- name: answer_text
dtype: string
- name: source
dtype: string
- name: images
list: string
splits:
- name: train
num_bytes: 224717958
num_examples: 258268
download_size: 39590089
dataset_size: 224717958
configs:
- config_name: default
data_files:
- split: train
path: data/train-*