https://absentbornlabs.org/writing/ Notes from running big models on small hardware. Measured results, including the things that did not work. en Thu, 20 Aug 2026 09:00:00 +0000 https://absentbornlabs.org/writing/model-from-scratch/ https://absentbornlabs.org/writing/model-from-scratch/ Thu, 20 Aug 2026 09:00:00 +0000 29.9M parameters, my own architecture code and my own tokenizer, random weights to coherent prose in sixteen minutes. Nine hours on one RTX 3070 Ti — including the two silent bugs that made the first attempt eighty-six times too slow. https://absentbornlabs.org/writing/27b-on-8gb/ https://absentbornlabs.org/writing/27b-on-8gb/ Thu, 20 Aug 2026 09:00:00 +0000 Six experiments on a single RTX 3070 Ti. Multi-token prediction doubled throughput for free, speculative decoding made things worse, and the biggest win turned out to be a Windows setting that was silently moving VRAM into system RAM.