{
  "context": {
    "cap_tokens": 129984,
    "max_model_len": 131072,
    "targets": []
  },
  "evaluation_protocol": {
    "budget_semantics": "visible_answer_tokens plus reasoning_headroom_tokens are sent as one max_tokens cap; the endpoint does not hard-partition the channels",
    "minimum_comparable_repetitions": 3,
    "minimum_pass_rate": 1.0,
    "reasoning_headroom_tokens": 0,
    "records_finish_reason": true,
    "records_full_visible_answer": true,
    "records_reasoning_channel_metadata": true,
    "repetitions": 2,
    "version": 3,
    "visible_answer_tokens": 512
  },
  "failures": [
    {
      "error": "completion_marker",
      "eval_id": "repeat_256_tokens",
      "failure_classes": [
        "visible_answer_budget_exhausted"
      ],
      "pass_rate": 0.0,
      "suite": "long-generation-256"
    }
  ],
  "identity": {
    "base_url": "http://127.0.0.1:39038/v1",
    "candidate_id": "google/gemma-4-31B-it-qat-w4a16-ct",
    "config_id": "gemma4-31b-qat-128k-c1-baseline-warm",
    "engine": "vllm-0.25.1",
    "gpu": "rtx-pro-6000-max-q",
    "model": "gemma4-heavy-lab",
    "started_at": "2026-07-17T04:30:15Z"
  },
  "intelligence": {
    "checks": [],
    "status": "not_run"
  },
  "run_id": "fast-bakeoff-20260717T043015Z",
  "schema": "anvil-serving.fast-tier-bakeoff/v1",
  "score_inputs": {
    "e2e_p50_ms": 0.0,
    "intelligence_pass_rate": null,
    "operational_fit_notes": [
      "endpoint was already loaded; benchmark did not start or stop serves"
    ],
    "session_recall_passed": false,
    "thinking_mode": "disabled",
    "tool_call_passed": false,
    "ttft_p50_ms": 0.0,
    "usable_context_tokens": null,
    "voice_latency_ms": null
  },
  "selection": {
    "context_targets": [
      32768
    ],
    "endpoint_already_loaded": true,
    "requests_per_context": 1,
    "suites": []
  },
  "session": {
    "checks": [],
    "status": "not_run"
  },
  "source_recipe": {
    "ref": null,
    "serve_command": null
  },
  "suites": {
    "long-generation-256": {
      "checks": [
        {
          "attempt_count": 2,
          "attempts": [
            {
              "attempt": 1,
              "budget": {
                "legacy_max_tokens_as_visible": false,
                "max_completion_tokens": 512,
                "reasoning_headroom_tokens": 0,
                "visible_answer_tokens": 512
              },
              "content": "x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x",
              "content_excerpt": "x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x ",
              "error": "completion_marker",
              "failure_class": "visible_answer_budget_exhausted",
              "finish_reason": "length",
              "latency_ms": 8270.382165908813,
              "reasoning_chars": 0,
              "reasoning_excerpt": "",
              "reasoning_field": null,
              "reasoning_tokens": null,
              "status": "failed",
              "text_checks": [
                {
                  "name": "completion_marker",
                  "passed": false
                }
              ],
              "usage": {
                "completion_tokens": 512,
                "prompt_tokens": 49,
                "prompt_tokens_details": null,
                "total_tokens": 561
              }
            },
            {
              "attempt": 2,
              "budget": {
                "legacy_max_tokens_as_visible": false,
                "max_completion_tokens": 512,
                "reasoning_headroom_tokens": 0,
                "visible_answer_tokens": 512
              },
              "content": "x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x",
              "content_excerpt": "x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x x ",
              "error": "completion_marker",
              "failure_class": "visible_answer_budget_exhausted",
              "finish_reason": "length",
              "latency_ms": 8171.473503112793,
              "reasoning_chars": 0,
              "reasoning_excerpt": "",
              "reasoning_field": null,
              "reasoning_tokens": null,
              "status": "failed",
              "text_checks": [
                {
                  "name": "completion_marker",
                  "passed": false
                }
              ],
              "usage": {
                "completion_tokens": 512,
                "prompt_tokens": 49,
                "prompt_tokens_details": null,
                "total_tokens": 561
              }
            }
          ],
          "error": "completion_marker",
          "id": "repeat_256_tokens",
          "pass_count": 0,
          "pass_rate": 0.0,
          "required_pass_rate": 1.0,
          "status": "failed",
          "validator": "deterministic_text_checks"
        }
      ],
      "date": "2026-07-17",
      "evidence_use": "diagnostic",
      "source": "docs/findings/2026-07-17-gemma4-31b-optimization-evidence/long-generation-256.suite.json",
      "source_sha256": "e02b91ec3d9220b5f1a0d2381aae0db550d4606d93e852fbacab11ca5e430106",
      "status": "failed",
      "validator_strength": "deterministic_marker",
      "work_class": "serving-throughput"
    }
  },
  "thinking": {
    "chat_template_kwargs": {
      "enable_thinking": false
    },
    "control_evidence": null,
    "control_evidence_sha256": null,
    "control_mechanism": "chat_template_kwargs",
    "control_requested": {
      "enable_thinking": false
    },
    "control_status": "requested_unverified",
    "mode": "disabled",
    "reasoning_effort": null,
    "unsupported": false
  },
  "timing": {
    "chat": {
      "e2e_p50_ms": 0.0,
      "e2e_p95_ms": 0.0,
      "output_tokens": 0,
      "ttft_p50_ms": 0.0,
      "ttft_p95_ms": 0.0
    },
    "wall_ms": 16467.000722885132
  },
  "tool": {
    "checks": [],
    "status": "not_run"
  },
  "voice": {
    "llm_latency_ms": null,
    "status": "not_run",
    "stt_latency_ms": null,
    "total_turn_latency_ms": null,
    "tts_latency_ms": null
  }
}