News for hackers. Stuff that matters.

A comment on How Meta trains large language models at scale. Parent comment

Re:How Meta trains large language models at scale by 123yawaworht456 on Thursday June 13, 2024 @06:00AM (#40667752)
1. improving their adtech. someone else's API offerings are not an option due to the sheer volume, PII and whatnot.

2. virtually free moderation for their existing (facebook, instagram, threads) and future social media services. likewise, their volume is too insane to even consider paying someone else to process it.

the models they do release are probably toys in comparison to their internal models.

[ Parent | on HN ]
Re:How Meta trains large language models at scale by takumif on Friday June 14, 2024 @02:34AM (#40678098)
> the models they do release are probably toys in comparison to their internal models.

I'm not sure if that's true. They did say they're going to make Llama 3 400B public and if they follow through, that's no toy and I doubt they have something significantly better internally.

[ Parent | on HN ]