News for hackers. Stuff that matters.

A comment on We fine-tuned Llama 405B on AMD GPUs. Parent comment

Re:We fine-tuned Llama 405B on AMD GPUs by itsTyrion on Thursday September 26, 2024 @08:11AM (#41657420)
Damn, ML/AI performance is that different? In games, 4090 -> 7900 XTX is more like -20%
[ Parent | on HN ]

There are no comments on this comment yet.