Soumith Chintala on Twitter: "i might have heard the same 😃 -- I guess info like this is passed around but no one wants to say it out loud. GPT-4: 8 x 220B experts trained with different data/task distributions and 16-iter inference. Glad that Geohot said it out loud. Though, at this point, GPT-4 is…" / Twitter - discu.eu

Hacker News

GPT-4: 8 x 220B experts trained with different data/task distributions https://twitter.com/soumithchintala/status/1671267150101721090 5 comments 20/6/2023

Linking pages

GPT-4's Secret Has Been Revealed - by Alberto Romero https://thealgorithmicbridge.substack.com/p/gpt-4s-secret-has-been-revealed 62 comments
Code Interpreter == GPT 4.5 (w/ Simon Willison & Alex Volkov) https://www.latent.space/p/code-interpreter 4 comments
Knowing Enough About MoE to Explain Dropped Tokens in GPT-4 - 152334H https://152334h.github.io/blog/knowing-enough-about-moe/ 1 comment
Open Source AI is AI we can Trust — with Soumith Chintala of Meta AI https://www.latent.space/p/soumith 0 comments

Related searches:

Search whole site: site:twitter.com

Search title: Soumith Chintala on Twitter: "i might have heard the same 😃 -- I guess info like this is passed around but no one wants to say it out loud. GPT-4: 8 x 220B experts trained with different data/task distributions and 16-iter inference. Glad that Geohot said it out loud. Though, at this point, GPT-4 is…" / Twitter

See how to search.

Submit link to: