metadata
language:
- en
dataset_info:
features:
- name: title
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 1183929798
num_examples: 2110473
download_size: 680523539
dataset_size: 1183929798
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Text split version of level 5 vital wikipedia articles. Text has been split on \n\n, and any instances where the length of the text was < 10 words were removed to avoid headings.