Gemma 4 on AI Insights Labhttps://cskwork.github.io/tags/gemma-4/Recent content in Gemma 4 on AI Insights LabHugoen-usSat, 12 Sep 2026 08:37:08 +0900Tiny on-device LLMs in 2026: which sub-1B models are worth running on a Raspberry Pi, an old laptop, or a CPU-only boxhttps://cskwork.github.io/posts/tiny-on-device-llms-sub-1b-models-2026/Sat, 12 Sep 2026 08:10:00 +0900https://cskwork.github.io/posts/tiny-on-device-llms-sub-1b-models-2026/Six sub-1B to 2B models compared for low-spec hardware as of September 2026: Needle 2, FunctionGemma 270M, LFM2.5-350M, Qwen3 0.6B, Llama 3.2 1B, and Gemma 4 E2B. Memory, speed, what each is good at, what it cannot do, and a llama.cpp walkthrough. Numbers from official model cards, with sources.