Package a reproducible self-hosted SmolVLM2 bootstrap #10

New Issue

timmy · 2026-08-19T01:58:58Z

timmy commented

2026-08-19 01:58:58 +00:00

Parent epic: #2

Outcome

Replace machine-specific spike steps with a pinned, repeatable local worker bootstrap.

Acceptance criteria

llama.cpp commit and GGUF hashes are pinned.
Install/start/health/stop commands work from a clean host.
Worker binds privately by default and no model weights enter Git.

Dependencies

None

Required evidence

Test or executable receipt attached to the issue/PR
Privacy/safety boundary checked where applicable
Clean checkout and relevant full suite pass

Delivery

Open a focused branch and PR with Closes #ISSUE. Do not include real medical images in issue comments or Git history.

Parent epic: #2 ## Outcome Replace machine-specific spike steps with a pinned, repeatable local worker bootstrap. ## Acceptance criteria - [ ] llama.cpp commit and GGUF hashes are pinned. - [ ] Install/start/health/stop commands work from a clean host. - [ ] Worker binds privately by default and no model weights enter Git. ## Dependencies - None ## Required evidence - Test or executable receipt attached to the issue/PR - Privacy/safety boundary checked where applicable - Clean checkout and relevant full suite pass ## Delivery Open a focused branch and PR with `Closes #ISSUE`. Do not include real medical images in issue comments or Git history.

timmy added this to the M0 — Triage & Reproducible Baseline milestone 2026-08-19 01:58:58 +00:00

timmy added the

type/ops

priority/P0

area/model

labels 2026-08-19 01:58:58 +00:00

timmy referenced this issue

2026-08-19 01:58:59 +00:00

Containerize the llama.cpp GPU worker with pinned model receipts #15

timmy referenced this issue

2026-08-19 01:59:03 +00:00

EPIC: Sovereign vision inference #2