metadata
dataset_info:
features:
- name: passage_id
dtype: int64
- name: title
dtype: string
- name: section_title
dtype: string
- name: text
dtype: string
- name: entities
sequence: string
- name: ids
sequence: int64
- name: passage_ids
sequence: int64
splits:
- name: train
num_bytes: 5539748471
num_examples: 10219164
download_size: 2893546048
dataset_size: 5539748471
configs:
- config_name: default
data_files:
- split: train
path: data/train-*