metadata
dataset_info:
features:
- name: question
dtype: string
- name: answer
dtype: string
- name: evidence
dtype: string
- name: paper_id
dtype: string
- name: paper_sec
dtype: string
- name: eval_3
dtype: float64
splits:
- name: train
num_bytes: 67572
num_examples: 155
download_size: 35627
dataset_size: 67572
configs:
- config_name: default
data_files:
- split: train
path: data/train-*