Josh Siviour
f9c84919ef
* feat: co-served GPU model bundles (catalog, lifecycle, embed/rerank routing) Serve the generative LLM, embedder and reranker as one co-resident GPU bundle: EMBEDDING/RERANK catalog modalities + co-serve group, bundle-aware admission (one GPU claim sized to summed member VRAM), single-flight one-VM launch with a member->port EndpointRecord, DynamicEndpointResolver member routing. Ochrevector embed/rerank/token-introspection resolve through the resolver (static single-endpoint map retired); rerank degrades to plain KNN on absent/failure. Public InvokeModel-embeddings and top-level Rerank deferred. * feat: co-serve vLLM+TEI bundle image with per-engine readiness Replace the single-model vllm-serve unit with a generic bedrock-serve@ systemd template driven by per-member BEDROCK_* env files, dispatching to vLLM or text-embeddings-router by engine and binding each member to its own port and weights device. TEI serves directly on its serve port; readiness probes each member on the route its engine answers (vLLM /v1/models, TEI /health) with the engine family carried end-to-end from the catalog. * fix: serve nomic embedder correctly in the co-serve bundle Bound embeddings requests to the TEI client batch limit so a large document's chunks are split across requests instead of 422ing. Serve nomic with explicit mean pooling, since the flat weights layout omits 1_Pooling/config.json and TEI would otherwise default to CLS. Add bge-base-en-v1.5 as a standalone alternative embedder. * fix: make the bedrock reaper family-aware and pin-safe A pinned endpoint is now exempt from terminate-and-relaunch on scrape failure, mirroring how it is already exempt from idle reclaim, so a transient unreachability can no longer tear down operator- or commitment-managed serving. A bundle with no generative vLLM member exposes no vLLM load metrics, so it is liveness-probed on /health and kept warm rather than scraped for idleness on a route it does not serve and reaped in a loop. * test: lock in co-serve group-member convergence to one VM * fix: pin self-host HF repo/revision in catalog and stop nested config flattening Carry HFRepo/HFRevision per self-host catalog entry; weights pull defaults --hf-repo and --revision from them so a bare pull cannot silently grab a regressed upstream main (the nomic v5 config.json duplicate-field break). Restrict fixed config filenames to the repo root so a nested config.json (sentence-transformers 1_Pooling/) is dropped instead of flattening onto the model config, and refuse any two source paths that collide at one dest key. * fix: let endpoint delete clear an account-scoped pinned record and report honestly A bare 'endpoint delete' resolved the Global key, which never held a pinned, account-scoped record, then printed 'torn down' over the idempotent no-op while the record survived. Add --account to the delete command so an operator can target the account 'endpoint list' shows, and return DeleteEndpointOutput.Removed so the CLI names a real teardown versus a lookup that found nothing and points at --account. A gone VM already tolerates NotFound, so a scoped delete clears a record whose instance was terminated out of band.
Spinifex Documentation
This directory contains the Spinifex documentation. This is available in individual markdown files, and a generated website to aid navigation and browsing documentation at: https://docs.mulgadc.com
The AWS model operation coverage report is generated from Spinifex's dispatch tables and the pinned AWS API models. Run make aws-model-coverage to regenerate and print it.
Frontmatter
Each published page lives at <category>/<slug>/README.md and carries YAML frontmatter that the docs site turns into page metadata. Three fields have length rules, because search engines flag a title tag or meta description that falls outside them:
| Field | Renders as | Length |
|---|---|---|
title |
Page heading, sidebar label, landing card | Short — it is UI text |
seoTitle |
The browser and search-result title | 50-60 characters |
description |
The search-result snippet, not shown on the page | 150-160 characters |
seoTitle is used exactly as written and must end with — Spinifex Docs. Leave it out and the title falls back to title plus that suffix, which is normally too short to satisfy the rule.
Keep the two titles distinct. title is navigation text in a narrow sidebar, so stretching it to reach 50 characters damages the layout to satisfy a rule that only applies to the title tag. Put the search-facing wording in seoTitle.
A new page is only published once its path is added to scripts/docs-config.json in the mulga-docs repo. That repo's scripts/build-docs.js warns about any field outside its range.