Skip to content

Top Best Ask Show New Jobs

The Pile: An 800GB Dataset of Diverse Text for Language Modeling [pdf] (opens in new tab)

(pile.eleuther.ai)

1 pointsnixtaken5y ago0 comments

0 comments

1 comments · 1 top-level

dang5y ago

https://news.ycombinator.com/item?id=25607809

j / k navigate · click thread line to collapse