metadata
dataset_info:
features:
- name: id
dtype: int64
- name: passage
dtype: string
- name: metadata
struct:
- name: split
dtype: int64
- name: title
dtype: string
- name: token_count
dtype: int64
splits:
- name: train
num_bytes: 41455820
num_examples: 23039
download_size: 21018685
dataset_size: 41455820
configs:
- config_name: default
data_files:
- split: train
path: data/train-*