metadata
dataset_info:
features:
- name: atomic
dtype: string
- name: retrieval
list:
- name: text
dtype: string
- name: title
dtype: string
splits:
- name: train
num_bytes: 2144514789
num_examples: 399586
download_size: 245359219
dataset_size: 2144514789
configs:
- config_name: default
data_files:
- split: train
path: data/train-*