Skip to content

Dataset · common-pile

YouTube video descriptions under CC BY license

common-pile/youtube

Manually curated descriptions of over 2,000 Creative Commons licensed YouTube channels.

What's inside

Contains English text data ranging from 1 to 10 million examples. Hand-curated from over 2,000 YouTube channels consistently producing original content under the CC BY license to combat license spoofing.

What you can do with it

  • Pre-training language models on high-quality web texts
  • Researching open speech and text content from YouTube
  • Creating diverse text corpora
Tasks
Text generation
Languages
English
Size
1M–10M rows
Downloads this month
1.1K
Likes
14

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("common-pile/youtube")