These 183,000 Books Are Fueling the Biggest Fight in Publishing and Tech
www.theatlantic.com
Use our new search tool to see which authors have been used to train the machines.
  • akrot@lemmy.worldEnglish
    11·
    1 year ago

    For anyone interesred, books3 were part of The Pile data used to train LLMs. They used to be hosted by The Eye, but recently removed due to DMCA. Their torrent link is still up though.