Every story and comment from the Hacker News API, in a layout borrowed from 2001.
A comment on Ornith-1.5: From Self-Scaffolding to Self-Improvement. Parent comment
tbh, I have stopped using MoE in the name of speed, the dense (with more active parameters) makes a real difference in output quality
There are no comments on this comment yet.
tbh, I have stopped using MoE in the name of speed, the dense (with more active parameters) makes a real difference in output quality