machine-learning
28 images, 800 steps, and a 49-minute SDXL face LoRA run on a 32 GB M2 Max. The dataset preparation, MPS settings, local privacy controls, and checks behind the finished adapter.
machine-learning
I swapped the Jev API for Laya, an open 421M decision model, in my Mario controller. Out of the box it died at the first Goomba. After a shorter prompt, two fine-tunes and a fallback, it cleared World 1-1 with 7 Jev calls instead of 208.
machine-learning
Jev could clear World 1-1, but every move was a 600 ms API call. I copied it into a local model, then trained a PPO agent from raw pixels. It cleared the level 10 times out of 10, then got worse.
debugging
I connected Jev to Super Mario Bros. and watched it repeat the same mistakes. Fixing timing, jump inputs, and experience memory eventually produced seven recorded World 1-1 clears.