// 1 transmissions tagged with #m4
A developer ran local models on an M4 instance with 24GB memory, demonstrating its feasibility for development [hn-front].
