noonghunnaandClaude Opus 4.8
d080757c2d
Surfaced by the gemma-4-12b registry handoff (Codex review): - Registry slug prefix != filesystem dir: slugs are vllm/llamacpp/ik-llama/beellama (llamacpp, NOT llama-cpp — registry-emit special-cases it). Fixes the slug drift. - Step 3 engine list: add ik-llama/beellama; ModelProfile engine enum: add beellama. - ModelProfile schema block flagged ILLUSTRATIVE (family-specific) → point at real current profiles (gemma-4-12b/-31b dense-SWA fields, qwen hybrid, MoE) not a flat head_dim/attention_type skeleton. - _entry sample: add status=/status_note= + emoji↔status mapping; llama.cpp-family note (drafter=None, kvcalc_key=SKIP); 'Registry entry != default' warning (no DEFAULTS row for experimental/preview/upstream-gated). - Step 4b: broaden vendored-CHAT-TEMPLATE guidance → vendored RUNTIME overlays / install_scripts (patches.yml + engine vendored_overlays + compose mount/entrypoint + diagnose hint; e.g. the gemma4 p-RoPE fix). - Guard caveats: test-compose-registry-disk globs disk (untracked scratch composes fail it → git ls-files preflight); test-switch-registry-parity engine allowlist. - Step 6: CalibrationData is vLLM-only (SKIP-key llama.cpp-family entries → BENCHMARKS + learnings, no kv-calc anchors). - Typo: bash → python3 tools/kv-calc.py. Doc-only; no registry/code touched (those are Codex's in-flight WIP). Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>
club-3090 docs index
Two tracks. Pick the one that matches what you're doing.
- User track — "I have GPUs and a model; how do I serve it?"
- Contributor / maintainer track — "I'm working on the v0.8.0 pull pipeline, patches, or the calibration loop."
Every link below resolves to a file in this repo.
User track
Start here if you want to run a model.
| Doc | What it is |
|---|---|
GETTING_STARTED.md |
Start here — 5-minute clone-to-curl path. No decisions, no menus. |
SINGLE_CARD.md |
1× RTX 3090 — workload → curated config → quick start. |
DUAL_CARD.md |
2× RTX 3090 (PCIe / NVLink auto-detected) — workload → config → quick start. |
MULTI_CARD.md |
3+ GPUs — TP scaling math, derivation from dual.yml, valid TP values. |
PULL.md |
Any HF safetensors repo — evaluate against the KV math, honest about confidence. |
BRING_YOUR_OWN.md |
Serve + tune + validate your own model/compose (any engine, single or dual) without touching the catalog. |
HARDWARE.md |
Card-class questions — 4090/5090, power caps, NVLink, laptop EC power. |
GLOSSARY.md |
TPS / KV / MTP / TP and the rest of the vocabulary. |
FAQ.md |
Common setup and operational questions. |
COMPARISONS.md |
Self-host vs cloud APIs — cost crossover and when each wins. |
EXAMPLES.md |
Worked end-to-end usage examples. |
Contributor / maintainer track
The v0.8.0 pull pipeline (in pipeline order)
A model slug flows through these stages. Read them in order to understand the whole.
| Stage | Doc | What it owns |
|---|---|---|
[D] |
COMPOSE_GENERATOR.md |
The #141 compose generator — the substrate that owns the arch→patches matrix. |
| Gate | PULL_GATE.md |
scripts/pull.sh — the locked 6-stratum abort taxonomy, [C0]/[C2a]/[B]/[C1] gates, §4.1 confidence×verdict table. |
[E] |
PULL_EMIT_DERIVED.md |
Download → boot → smoke for a download-eligible derived model; writes the §6 capture artifacts. |
[F] |
LOOP.md |
The calibration loop — reads the capture bundle, classifies, runs the inbound-trust pipeline, dedups failures into the tracker. |
Patch & model contribution
| Doc | What it is |
|---|---|
PATCH_POLICY.md |
When/how a patch ships, the local-overlay vs upstream rules. |
PATCH_ATTRIBUTION.md |
The Phase-A patch-attribution matrix — arch → engine-pin → required patches. |
ADDING_MODELS.md |
How a new model gets added to the curated catalog. |
KV_MATH.md |
The KV-cache math the [B] fit verdict is computed from. |
Stack reference & ops
| Doc | What it is |
|---|---|
ARCHITECTURE.md |
Repo/stack architecture overview. |
UPSTREAM.md |
Upstream PR / issue tracker for this stack. |
NIGHTLY_BUMP_RUNBOOK.md |
Procedure for bumping the vLLM nightly pin. |
CONTAINER_RUNTIMES.md |
Docker / container runtime notes. |
Reference matrices & deep dives
These are cross-cutting references both tracks reach for.
| Doc | What it is |
|---|---|
scripts/switch.sh --list (runtime command, not a doc) |
The authoritative compose × slug matrix. Registry-derived from scripts/lib/profiles/compose_registry.py, so it's always current — every launchable slug with its topology, model, engine, KV format, and max ctx. Run this rather than trusting any hand-maintained table; the static lists in the per-topology docs are illustrative, this is the source of truth. |
engines/ |
Per-engine deep dives — vLLM, llama.cpp, SGLang. |
INFERENCE_ENGINES.md |
Engine picker — which engine for which workload, and structural gaps. |
CLIFFS.md |
The accumulated-context / prefill failure modes (Cliff 2, Cliff 2b) and how to detect them. |
DTYPE_MATRIX.md |
Supported dtype × model × engine matrix. |
KERNEL_MATRIX.md |
Quant-kernel availability and alignment constraints. |
QUALITY_TEST.md |
The quality-test harness and what it measures. |
RESULTS_CARD.md |
The standard 3-panel format (Serving · Quality · Takeaways) for sharing a config's measured results. |
STRUCTURED_COT.md |
The bounded-thinking / structured-CoT compose path. |
TQ3_MTP_GENESIS.md |
TQ3 KV × MTP × Genesis-patch results and config. |