---
name: radiance-w4-campaign
description: "Run a radiance GPTQ-w4 quantization campaign on dankserver: calibration → protected experts → convert → KLD gate → deploy, with the engine-pin and MNBT laws."
---

Pipeline for quantizing a BF16 Qwen3.8-Flash-Next checkpoint into a servable `qwen4exp` `.rad` on dankserver (frank-primary). Reference run: `serverAppConfigs-docs/LLM/frank-primary-radiance-swift15-w4dn8.md`; scripts in `scripts/` (`run-swift15-bf16.sh`, `calib-push.py`, `derive-protected-experts.py`, `apply-protected-cast.sh`, `convert-swift15-w4dn8.sh`, `kld-gate-swift15.sh`, `fleet-probes-swift15.sh`).

Sequence:
1. **Fetch** checkpoint with per-shard LFS sha256 verification (`-f`, byte counts, no token on argv).
2. **BF16 container**: `rad-convert` no-recipe (pure copy) — serves as the KLD reference engine.
3. **KLD reference**: record logits on the SERVE pin (1.0.13) at MNBT 32768, `--max-model-len 16384`, corpus ≤16k prompts. Writes atomically at completion (`rows:` line = receipt). Never scp a running bash script (lazy-read law).
4. **Calibration**: engine **1.2.0** (only pin with `RADIANCE_CALIB_TOKENS`); push ~10M tokens of the model's own distribution; per-expert down-proj Grams land as RADGRAM2 files (`gram.r<rank>.blk.L.ffn_down_exps.<e>.bin`). Grams are engine-invariant across pins — the version boundary is by design.
5. **Protected experts**: derive from gram energy (share ≥3 % AND ≥3× routing share), splice between `# PROTECTED-BEGIN/END` markers of the recipe. Plugin law: ONE encoding per expert table; only whole bf16 experts legal beside it — no gate_up/down split.
6. **Final convert** on the SERVE pin (CPU GPTQ, hours; disk gate ≥146 GiB free first).
7. **KLD gate** on the serve pin: **MNBT ≤ 2048** (w4 streams in place only there; 32768 → `DID NOT FIT ... stored form is not the container's`), pool 40960, fp8 KV matching the production KV dtype. Gate: mean KL ≤ 0.03, top-1 ≥ 95 %.
8. **Deploy** `./scripts/deploy.sh dankserver --service frank-primary`; verify `/v1/models` id = served-model-name (alias independent of file — harness models.yml untouched), boot ledger, streamed ≥128-token decode probe, fleet probes (vision needs ≥256-token budget + `reasoning_content` fallback).

Laws: stop frank-primary for every GPU phase (relinked unit: stop, mask fails); verify docs against `podman inspect <name> --format {{join .Config.Cmd " "}}` (rootless, as dank) before writing shapes; ff-pull fails on untracked files the merge tracks — trash first.