Sources
Hubs, tools, Norwegian models, a memory filter and a GPU overview for local AI.
Norwegian reference for local AI — registers, plans and honest tests.
kjerneki collects registers, build plans and honest tests for people who want to run AI on their own machine — without the cloud when that is possible.
Hubs, tools, Norwegian models, a memory filter and a GPU overview for local AI.
Short advice for local AI — runtime, model cards, memory and workflow.
Longer tests and notes. Start with Hermes vs Hyperagent.
Short explanations of common model licenses, with links to the official text. Not legal advice.
Seed register of models that fit local runs. RAM and our own tests only where we have a source or have tried them.
Quantization, GGUF, VRAM, Ollama and other words you meet when you build local AI.
Start with Jetson Orin Nano Super — a skeleton with a parts list and steps marked as draft.
Two tests under Tests: faster-whisper (run, 0 NOK) and Surface Book 2 / HyperFrames (planned, not tested yet).
Want to suggest a model, license row or plan? The suggestion channel is coming. For now: see suggestions.
New tests, registers and plans for local AI. Free — unsubscribe any time. (Form labels are in Norwegian.)