Hacker News new | ask | show | jobs
by lossolo 51 days ago
It's fine tuned Kimi, they didn't train it from scratch.
1 comments
YmiYugy 51 days ago
Sure, but so what? It seems that model size and RL are the determining factors these days.
link