Dev trains continually-learning mini-AGI from scratch on an 8GB laptop
A developer open-sourced mini-AGI, a continually learning byte-level language model trained from scratch on an 8GB RTX 3070 laptop, now at 530M parameters with a growing pool of 169 experts; full weights are due in weeks.
2026-09-21 ~ 2026-09-23 · 4 related posts
- Dev Trains Continual-Learning Model From Scratch on 8GB VRAM Laptop, Now at 530M Params — Another__one · 2026-09-21
- Show HN: Mini-AGI — a dynamic continual learning model trained on 8GB VRAM — petrusenko_max · 2026-09-22
- mini-AGI: Looped Continual-Learning Transformer With Evolutionary Experts Trained on a Laptop — returnity · 2026-09-23
- mini-AGI: continual learning model trained from scratch on an 8GB VRAM laptop GPU — udmrzn · 2026-09-23