metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: answer
sequence: string
- name: alt_answers
sequence: string
- name: evidence_sentences
sequence: string
splits:
- name: train
num_bytes: 10549593
num_examples: 4927
download_size: 5395777
dataset_size: 10549593
configs:
- config_name: default
data_files:
- split: train
path: data/train-*