3 ms·
This sub thread isn’t about hosted inference though. I interpreted the question as “is there a reasonable laptop out there that can run large models locally in
by adastra22 3y ago
This sub thread isn’t about hosted inference though. I interpreted the question as “is there a reasonable laptop out there that can run large models locally in real time?” And a MacBook Air does fit the bill for that.
It’s an entry-level Mac, and it is capable of doing inference jobs. If you have around $1k available to spend on a daily driver laptop, but you also want to do some AI inference experiments, the MacBook Air is what I’d recommend.
The AMD Framework laptop might be a good alternative. But I don’t know how well its integrated GPU is supported by llama.cpp.