Same machine, same Python, same one-line command. The result depended on who launched that Python. When I opened an SSH session by hand and typed python -c “import torch”, it passed. When my daemon brought the GPU image server up on that machine automatically, the same import torch died like this: OSError: [WinError 4551] ……
I built a loop that puts up a personal homepage on its own. It’s a static site, résumé-shaped, and I cut the work into units. Every unit spawns a fresh cloud agent. That agent plans, and it doesn’t write the code itself. It hands the code off to a local model on a remote machine….
My GPU worked for 28 minutes and 37 seconds without a single complaint. Every step ran, and sampling finished. Then the last node in the graph printed this. RuntimeError: expected input to have 48 channels, but got 16 Decoding never started. Everything those 28 minutes and 37 seconds produced went straight in the bin. The…
My development rules contain one line: Code generation goes to the local model. The cloud model plans and reviews. On 26 June 2026 I added two words to it: no exceptions. That phrase survived five days. On 1 July I conceded the first exception and rewrote the line myself. The rule itself came out of…
I own four keyboards. Different switches, different sizes, bought years apart. They all run one layout. Every one of them is set up the way my HHKB is set up: the Caps Lock position produces Ctrl or Command, and everything else matches as far as each board allows. The HHKB is the reference, and the…
I built a thing that answers when I talk to it. It takes the microphone, turns what I said into text, hands that to a model, and speaks the reply back out loud. I never touched a cloud speech API. The brain in the middle is a different story, and I deal with it head…
My coding sessions run on a machine at home, and I’m away from that machine far more than I’m sitting in front of it. So I built a way to keep typing into those same sessions from my phone. Moonlight does this for game screens. I did it for text. The most expensive thing that…
Two local embedding models ranked the same 24 bookmarks identically — 8 of 8 correct each. Their similarity scores sat in almost non-overlapping ranges, and one hardcoded 0.75 cutoff silently erased every result from one of them.
A cloud model wrote 1,900 lines in one session and burned my five-hour limit — then the next session remembered none of it. Three layers of context files plus hooks, and what each one is actually for.
Fitting a 30B MoE model into 16GB of VRAM comes down to one flag. Full n-cpu-moe sweeps on a 16GB and an 11GB card — including the silent VRAM overflow that costs 2.5x with no error.