docs: cookbook benchmark accuracy labels come from the model config (no engine default) (#27842)

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>

.claude/skills/cookbook-add-model/references/authoring-reference.md

@@ -47,7 +47,8 @@ the full contract):
 | `dockerImages` | `{[hwId]: string}` | Per-hw image name for `docker run` framing. **Ask the user which sglang build the recipes ran on; don't guess a supporting release.** Falls back to `lmsysorg/sglang:dev` if missing — also the sensible default when unsure. |
 | `playgroundFeatures` | `{[axisId]: {...}}` | Opts into the Playground widget. See §2.3. |
 | `benchmarkCommands` | `{speed: string, accuracy: {[accKey]: string \| {[variant]: string}}, numPromptsByConc?: {[c]: number}}` | Powers the benchmark card's **"⚡ Reproduce"** modal. `speed` is ONE `bench_serving` template; the engine fills `{{DATASET}}`/`{{ISL}}`/`{{OSL}}` from each cell's `speed[].workload`, the chip-picked `{{MAX_CONCURRENCY}}`, and `{{NUM_PROMPTS}}` (resolved `workload.num_prompts ?? numPromptsByConc[c] ?? max(c*2, 200)`). `accuracy` maps an accuracy field (e.g. `gsm8k_pct`) to a per-eval template — a string, OR a `{flash, pro, …}` object keyed by variant when the command differs per variant (e.g. GPQA/AIME `--max-tokens`). The modal renders a chip per eval (one command area, like Speed). Both also use `{{MODEL_NAME}}` + `{{CURL_HOST}}`/`{{CURL_PORT}}` like `curl`. Optional; the button only appears when this AND `benchmarks` are present. |
-| `defaultAccuracy` | `{[variant]: {[accKey]: number}}` | Model-level accuracy applied to **every** cell of a variant (e.g. GPQA Diamond / AIME25 — hardware-independent). Merged UNDER each cell's measured `accuracy` (a per-cell value wins), so you set a variant's score once instead of copying it onto every benchmark entry. Keys must match `ACCURACY_LABELS` + `benchmarkCommands.accuracy`. |
+| `defaultAccuracy` | `{[variant]: {[accKey]: number}}` | Model-level accuracy applied to **every** cell of a variant (e.g. GPQA Diamond / AIME25 — hardware-independent). Merged UNDER each cell's measured `accuracy` (a per-cell value wins), so you set a variant's score once instead of copying it onto every benchmark entry. Keys must match `accuracyLabels` (below) + `benchmarkCommands.accuracy`. |
+| `accuracyLabels` | `[key, label, unit][]` | The eval set rendered in the benchmark card and the "⚡ Reproduce" modal — **the engine ships no default**, every config declares its own (e.g. DSv4: GPQA/AIME25/GSM8K; Qwen3.5: GSM8K/MMMU). Required whenever the benchmarks carry accuracy data; without it the accuracy rows silently don't render. Every key used in `benchmarks[].accuracy`, `defaultAccuracy`, and `benchmarkCommands.accuracy` must appear here. |
 | `github` | `{owner?, repo?, issueTemplate?, cookbookModel?}` | Overrides for the "Submit verified cell" CTA in the playground. Defaults: `sgl-project/sglang` + `3-playground-verified-cell.yml` + `"deepseek-ai/deepseek-v4"`. Set `cookbookModel` to the model's HF id (`<hf-org>/<model-slug>`); it prefills the issue template's free-form `model` input when the issue opens. **Don't prune this block** — without it the engine falls back to `deepseek-ai/deepseek-v4` and submissions from your page get mislabeled. |
 
 ## 2.2 Author the 5-dim matrix (`cells[]`)

.claude/skills/cookbook-add-model/templates/config.jsx.tmpl

@@ -108,11 +108,20 @@ sgl-eval run gsm8k \\
 
   // OPTIONAL — per-variant accuracy applied to EVERY cell of a variant (hardware-
   // independent, e.g. GPQA/AIME). Per-cell `accuracy` overrides. Keys must match
-  // ACCURACY_LABELS + benchmarkCommands.accuracy. Delete if no numbers yet.
+  // the effective accuracy labels + benchmarkCommands.accuracy. Delete if no numbers yet.
   defaultAccuracy: {
     default: { gsm8k_pct: null },
   },
 
+  // The eval set rendered in the benchmark card + "⚡ Reproduce" — the engine
+  // ships NO default; required whenever the benchmarks carry accuracy data
+  // (without it the accuracy rows silently don't render). [key, label, unit]
+  // tuples; keys must match benchmarks[].accuracy + defaultAccuracy +
+  // benchmarkCommands.accuracy. Delete only if there are no accuracy numbers.
+  accuracyLabels: [
+    ["gsm8k_pct", "GSM8K", "%"],
+  ],
+
   // OPTIONAL — `# ...` hint lines prepended to multi-node commands, ONLY for hw
   // whose fabric needs manual NIC env (e.g. gb200 NVL72/MNNVL). NOT every multi-N
   // hw needs this — standard-IB DeepEP / Marlin multi-node don't. Delete if unused.

.claude/skills/cookbook-review-pr/SKILL.md

@@ -78,7 +78,10 @@ than restating.
 
 ### 4. Benchmarks
 - Each `benchmarks[]` entry's `match` tuple corresponds to a real cell.
-- `defaultAccuracy` keys ∈ `ACCURACY_LABELS` (and `benchmarkCommands.accuracy`).
+- `accuracyLabels` is present whenever the benchmarks carry accuracy data — the engine
+  ships NO default eval set; without it the accuracy rows silently don't render.
+  `defaultAccuracy` / per-cell `accuracy` / `benchmarkCommands.accuracy` keys all
+  ∈ `config.accuracyLabels`.
 - A benchmark's quantization must match a variant actually listed — `(BF16)` on a model
   that only released FP8/FP4 is a factual bug.
 - `benchmarkCommands.speed` is `python3 -m sglang.bench_serving` (the workload), separate

docs_new/src/snippets/_deployment.jsx

@@ -20,6 +20,9 @@
 //   benchmarkCommands  optional — powers the "⚡ Reproduce" modal (speed +
 //                      per-eval accuracy templates)
 //   defaultAccuracy    optional — per-variant accuracy merged under cell.accuracy
+//   accuracyLabels     [key, label, unit][] — the eval set shown in the
+//                      benchmark card + "⚡ Reproduce". NO engine default:
+//                      required whenever benchmarks carry accuracy data
 //   multiNodeHints     optional — {[hwId]: string[]} prepended as `# ...` lines
 //   dockerImages       optional — per-hw image for `docker run` mode
 //   github             optional — "Submit verified cell" issue-template overrides
@@ -530,13 +533,12 @@ export const Deployment = ({ config, benchmarks }) => {
     return cmd;
   };
 
-  // Accuracy labels: [field-key, display-label, unit]. Keys must match the
-  // `accuracy` fields in the benchmarks file + `benchmarkCommands.accuracy`.
-  const ACCURACY_LABELS = [
-    ["gpqa_pct",   "GPQA Diamond", "%"],
-    ["aime25_pct", "AIME25",       "%"],
-    ["gsm8k_pct",  "GSM8K (1-shot)", "%"],
-  ];
+  // Accuracy labels: [field-key, display-label, unit]. Declared per model via
+  // `config.accuracyLabels` — the engine ships NO default eval set. A config
+  // without it renders no accuracy rows (and no Accuracy section in the
+  // "⚡ Reproduce" modal). Keys must match the `accuracy` fields in the
+  // benchmarks file + `benchmarkCommands.accuracy`.
+  const ACCURACY_LABELS = config.accuracyLabels || [];
 
   const renderBenchmarkCard = (entry) => {
     // [key, label, unit, compute?]. Optional compute(measurement) supplies

docs_new/src/snippets/configs/deepseek-ai/deepseek-v4.jsx

@@ -122,6 +122,14 @@ sgl-eval run aime25 \\
     pro:   { gpqa_pct: 90.1, aime25_pct: 97.5, gsm8k_pct: 96.13 },
   },
 
+  // The eval set rendered in the benchmark card + "⚡ Reproduce" (the engine
+  // ships no default — every config declares its own).
+  accuracyLabels: [
+    ["gpqa_pct",   "GPQA Diamond",   "%"],
+    ["aime25_pct", "AIME25",         "%"],
+    ["gsm8k_pct",  "GSM8K (1-shot)", "%"],
+  ],
+
   // Prepended as `# ...` comments above multi-node commands.
   multiNodeHints: {
     gb200: [