Jekyll2026-07-20T16:45:53+00:00https://anmolchaure.github.io/feed.xmlAnmol ChaureResearch Technician for Environmental Health Research Group, CULS, PragueAnmol ChaureRunning a Local LLM with Ollama2026-07-20T00:00:00+00:002026-07-20T00:00:00+00:00https://anmolchaure.github.io/notes/setting-up-ollamaWhy run an LLM locally

Running a model locally gives you control, privacy, and offline operationno per-token cost and no data leaving the device.

Hardware constraints

I have a CPU-only machine with 6GB RAM, so it will be best to stick to 3B-parameter models. A 3B model at Q4 quantization is roughly 2GB, which leaves headroom for my operating system.

Installing Ollama on Ubuntu

curl -fsSL https://ollama.com/install.sh | sh
ollama run llama3.2:3b
]]>Anmol Chaure