It uses Accelerate so it may be using some of that indirectly. | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		astrange on March 10, 2023 \| parent \| context \| favorite \| on: Llama.cpp: Port of Facebook's LLaMA model in C/C++... It uses Accelerate so it may be using some of that indirectly.

olliej on March 11, 2023 [–]

Maybe, but as far as I can tell using the ML specific hardware requires CoreML and CoreML's data formats.

But I also can't tell where the vector (or matrix?) extensions end and the apple "neural" engine begins so shrug? :D

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact