File size: 379 Bytes
f485f49 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 | dataset:
name: Mixed Text Dataset
total_samples: 20000
sources:
- name: Wikipedia
samples: 10000
details: "20220301 dump"
- name: Gutenberg
samples: 3000
details: "GutenDex API"
- name: CNN/DailyMail
samples: 7000
details: "Version 3.0.0"
preprocessing:
method: regex_cleaning
description: "Replaced patterns like (.*?/) with whitespace" |