logoalt Hacker News

lukeschlather • today at 5:23 PM • 3 replies • view on HN

Yeah, why didn't I think of just spending $30k/month on an inference server and figuring out how to train a 2T parameter model. I clearly don't care about this at all or I would just do that.


Replies

vorticalbox • today at 7:19 PM

Deep seek have released their inference engine.

https://github.com/deepseek-ai/DeepInfer

nozzlegear • today at 5:42 PM

> and figuring out how to train a 2T parameter model

Why train your own model? You could use one of the many models already available on HF.

ndnxjxjdj • today at 5:55 PM

literally you can just run GLM

➕ show 1 reply