Simpler (and faster): a single RTX Pro 6000 :)

embedding-shape • last Sunday at 10:14 AM • 2 replies • view on HN

androiddrew • last Sunday at 12:00 PM

Curious, do you find the 3.5 120B sized MoE works better than the dense 3.6 27B?

➕ show 1 reply

nsbk • last Sunday at 10:35 AM

Hahaha well played! Hat tip

alt Hacker News