Re:Meta Llama 3
by cjbprime on Thursday April 18, 2024 @01:18PM
(#40078511)
(You can't compare parameter count with a mixture of experts model, which is what the 1.8T rumor says that GPT-4 is.)
[ Parent
| on HN ]
You absolutely can since it has a size advantage either way. MoE means the expert model performs better BECAUSE of the overall model size.