6,294
edits
Meta Platforms:AI Capabilities: Difference between revisions
→Competitiveness
| Line 3: | Line 3: | ||
* Meta claims most humans like their Llama 3 more than competitive models.<ref>https://ai.meta.com/blog/meta-llama-3/</ref> | * Meta claims most humans like their Llama 3 more than competitive models.<ref>https://ai.meta.com/blog/meta-llama-3/</ref> | ||
* The company says its largest models, which are still in training are over 400B parameters. | * The company says its largest models, which are still in training are over 400B parameters. | ||
* Meta says their human evaluations was based on 1,800 prompts across 12 key uses<ref>https://www.youtube.com/watch?v=r3DC_gjFCSA</ref>. | |||
[[File:Llama 3.png|left|thumb|490x490px|Source: [https://ai.meta.com/blog/meta-llama-3/ Meta Platforms]]] | [[File:Llama 3.png|left|thumb|490x490px|Source: [https://ai.meta.com/blog/meta-llama-3/ Meta Platforms]]] | ||
| Line 34: | Line 35: | ||
[[File:Evolution of Llama ...png|left|thumb|Source: Meta]] | [[File:Evolution of Llama ...png|left|thumb|Source: Meta]] | ||