#retrocomputing

A high-signal LocalLLaMA post described a port of llama2.c to classic Mac OS that runs Karpathy’s TinyStories 260K model on a stock iMac G3. The project is compelling because most of the work is systems engineering: endianness fixes, memory partition management, and layout debugging on vintage hardware.

Hacker News surfaced ATTN/11, a project that trains a single-layer, single-head Transformer in PDP-11 assembly on a PDP-11/34A. The README says careful fixed-point math, per-layer learning rates, and a 32KB memory budget cut training from multi-hour estimates to a 5.5-minute run that reaches 10/10 accuracy on digit reversal.