Hacker News new | ask | show | jobs
by mathgeek 30 days ago
Good summary blog: https://maloyan.xyz/blog/running-qwen-locally-mac-mini-m4
1 comments

> That's not hypothetical — it's a real measurement on the base model Mac Mini.

Hmmm

Yes, I tried out that blog post and concluded that it has llm-written formulations, weird ordering of its alternatives, and LM Studio is the "Step" in its list that works best. The mmap trick certainly does not work.