DEDUPED Hot AI repos, turned into use cases Get 5 a day

How to boot a local LLM on a Raspberry Pi 5 from SD with NIGHTRUN

Build a Pi 5 NightRun image with a Llama or Granite model and boot it from a microSD card.

Who it is forA maker with a Raspberry Pi 5 who wants a self-contained offline chat device and can build pinned firmware once.

The problemA Pi normally runs a full Linux system for a local model, and the setup is heavy for a single-purpose device.

The outcomeA microSD card that boots the Pi 5 into a NightRun chat prompt with the chosen model in RAM.

What you need first

  • A Linux machine with Rust stable and nightly installed
  • About 6 GB of free disk for a model plus the image
  • A USB stick (x86_64 target) or microSD card (Raspberry Pi 5 target), or QEMU for testing without hardware
  • For real hardware: a 64-bit UEFI x86_64 PC with Secure Boot off, or a Raspberry Pi 5

Steps

  1. Run git clone https://github.com/hardrave/NIGHTRUN.git and then cd NIGHTRUN.
  2. Run scripts/build-rpi5-firmware.sh once to build the UEFI firmware from pinned source, following docs/rpi5-uefi.md.
  3. Run cargo xtask pi-image --model models/model.nrm to build nightrun-pi5.img.
  4. Run sudo dd if=nightrun-pi5.img of=/dev/REPLACE_WITH_SD_DEVICE bs=4M status=progress conv=fsync, replacing the placeholder with the whole microSD device after checking it with lsblk -d -o NAME,SIZE,MODEL,TRAN. (inferred, check the README)
  5. Insert the microSD card into the Pi 5 and power it on.

Prompt for your agent

Use the repo hardrave/NIGHTRUN to build a Raspberry Pi 5 image at models/model.nrm for the target <target> (a whole-disk microSD device). Expected output: nightrun-pi5.img built with cargo xtask pi-image and written to the card. Verify that the device name matches <target> before any write step runs.

Caveat. The README says the Pi 5 build was validated on a real D0-stepping 8 GB board, and the Llama 3.2 1B and Granite models need at least 4 GB and 6 GB of RAM respectively.

Share: X LinkedIn Hacker News Email Copy link

Disclaimer

Deduped lists open-source repositories that belong to other people. We do not own, maintain, audit, or vouch for any of them, and a repo appearing here is not an endorsement or a safety check. The use cases, steps, and prompts are written by AI and may be incomplete, outdated, or just wrong. Where a use case is marked "Seen in the wild", we link to public posts where someone described something similar, but we have not verified those posts and the use case itself is still our AI's own write-up. Use all of this at your own risk. Only run code, commands, or agents on systems and accounts you are authorized to use, and read the code and every command before you run it, especially anything that touches credentials, files, or the network. Nothing here is professional, security, legal, or financial advice. All trademarks and project names belong to their respective owners, and their use here does not imply affiliation. Content is provided as is, without warranties of any kind. Found a mistake, a safety problem, or something you want removed? Email [email protected] and we will look at it promptly. Full disclaimer page