metadata
dataset_info:
features:
- name: pubmed_ids
dtype: string
- name: questions
dtype: string
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 4303770
num_examples: 4719
download_size: 1657822
dataset_size: 4303770
configs:
- config_name: default
data_files:
- split: train
path: data/train-*