TSCCspecific-silver
Running Pi with local models
i've been running pi with GLM-4.7-Flash for a few weeks, so far pretty satisfied, although it's not as good as cloud, it can do some simple stuff pretty reliably. running on apple silicon, m1 max with 64gb ram (btw. 5 year old hardware) and it's important to use the raw mlx-lm inference engine with it, as it's the most performant
