Arthur Besse@lemmy.ml to Technology@beehaw.orgEnglish · edit-21 年前Sarah Silverman and other authors are suing OpenAI and Meta for copyright infringement, alleging that they're training their LLMs on books via Library Genesis and Z-Librarywww.thedailybeast.comexternal-linkmessage-square131fedilinkarrow-up1211arrow-down10cross-posted to: [email protected][email protected]
arrow-up1211arrow-down1external-linkSarah Silverman and other authors are suing OpenAI and Meta for copyright infringement, alleging that they're training their LLMs on books via Library Genesis and Z-Librarywww.thedailybeast.comArthur Besse@lemmy.ml to Technology@beehaw.orgEnglish · edit-21 年前message-square131fedilinkcross-posted to: [email protected][email protected]
minus-squareISMETAlinkfedilinkEnglisharrow-up2·1 年前GPT3 is 800GB while the entirety of the English Wikipedia is around 10GB compressed. So yeah it doesn’t store evey detail of everything but LLMs do memorize a lot of things verbatim. Also see https://bair.berkeley.edu/blog/2020/12/20/lmmem/
GPT3 is 800GB while the entirety of the English Wikipedia is around 10GB compressed. So yeah it doesn’t store evey detail of everything but LLMs do memorize a lot of things verbatim. Also see https://bair.berkeley.edu/blog/2020/12/20/lmmem/