metadata
dataset_info:
features:
- name: passage_id
dtype: int64
- name: passage
dtype: large_string
- name: type
dtype: large_string
- name: question
dtype: large_string
- name: answer
dtype: large_string
- name: source
dtype: large_string
splits:
- name: train
num_bytes: 493045651
num_examples: 505681
download_size: 30168557
dataset_size: 493045651
configs:
- config_name: default
data_files:
- split: train
path: data/train-*