{ "context": { "cap_tokens": 31680, "max_model_len": 32768, "targets": [ { "attempted_context_tokens": 8192, "chars_per_token": 3.0, "clamped_tokens": 8192, "e2e_ms": 1350.9016036987305, "estimated_prompt_tokens": 8192, "output_tokens": 27, "status": "passed", "target_tokens": 8192, "ttft_ms": 1220.9112644195557, "usage": { "completion_tokens": 28, "prompt_tokens": 6525, "total_tokens": 6553 } }, { "attempted_context_tokens": 24000, "chars_per_token": 3.766, "clamped_tokens": 24000, "e2e_ms": 1080.977439880371, "estimated_prompt_tokens": 23999, "output_tokens": 58, "status": "passed", "target_tokens": 24000, "ttft_ms": 744.2588806152344, "usage": { "completion_tokens": 60, "prompt_tokens": 25205, "total_tokens": 25265 } } ] }, "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": 3, "version": 3, "visible_answer_tokens": 512 }, "failures": [ { "error": "pass rate below threshold", "failure_classes": [ "deterministic_check_failed" ], "pass_rate": 0.3333333333333333, "prompt_id": "parallel_timeout_triage", "suite": "intelligence" } ], "identity": { "base_url": "http://127.0.0.1:39037/v1", "candidate_id": "unsloth-gemma4-26b-a4b-nvfp4", "config_id": "fast-32k-thinking-disabled", "engine": "vllm-0.25.1", "gpu": "RTX-5090-32GB", "model": "gemma4-fast-lab", "started_at": "2026-07-16T23:11:23Z" }, "intelligence": { "checks": [ { "attempt_count": 3, "attempts": [ { "attempt": 1, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "```diff\n--- app.py\n+++ app.py\n@@ -1,2 +1,2 @@\n-timeout = 30\n+timeout = 45\n retries = 2\n```", "content_excerpt": "```diff\n--- app.py\n+++ app.py\n@@ -1,2 +1,2 @@\n-timeout = 30\n+timeout = 45\n retries = 2\n```", "failure_class": null, "finish_reason": "stop", "latency_ms": 242.22803115844727, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "text_checks": [ { "name": "diff_shape", "passed": true }, { "name": "removes_old_timeout", "passed": true }, { "name": "adds_new_timeout", "passed": true } ], "usage": { "completion_tokens": 46, "prompt_tokens": 54, "prompt_tokens_details": null, "total_tokens": 100 } }, { "attempt": 2, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "```diff\n--- app.py\n+++ app.py\n@@ -1,2 +1,2 @@\n-timeout = 30\n+timeout = 45\n retries = 2\n```", "content_excerpt": "```diff\n--- app.py\n+++ app.py\n@@ -1,2 +1,2 @@\n-timeout = 30\n+timeout = 45\n retries = 2\n```", "failure_class": null, "finish_reason": "stop", "latency_ms": 250.1962184906006, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "text_checks": [ { "name": "diff_shape", "passed": true }, { "name": "removes_old_timeout", "passed": true }, { "name": "adds_new_timeout", "passed": true } ], "usage": { "completion_tokens": 46, "prompt_tokens": 54, "prompt_tokens_details": null, "total_tokens": 100 } }, { "attempt": 3, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "```diff\n--- app.py\n+++ app.py\n@@ -1,2 +1,2 @@\n-timeout = 30\n+timeout = 45\n retries = 2\n```", "content_excerpt": "```diff\n--- app.py\n+++ app.py\n@@ -1,2 +1,2 @@\n-timeout = 30\n+timeout = 45\n retries = 2\n```", "failure_class": null, "finish_reason": "stop", "latency_ms": 253.4956932067871, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "text_checks": [ { "name": "diff_shape", "passed": true }, { "name": "removes_old_timeout", "passed": true }, { "name": "adds_new_timeout", "passed": true } ], "usage": { "completion_tokens": 46, "prompt_tokens": 54, "prompt_tokens_details": null, "total_tokens": 100 } } ], "id": "unified_diff_timeout_edit", "pass_count": 3, "pass_rate": 1.0, "required_pass_rate": 1.0, "status": "passed", "validator": "deterministic_text_checks" }, { "attempt_count": 3, "attempts": [ { "attempt": 1, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "The total latency (3000 ms) exceeds the 2500 ms timeout due to high LLM processing time; implement **streaming** (sending LLM tokens to TTS immediately as they are generated) to reduce perceived and total latency.", "content_excerpt": "The total latency (3000 ms) exceeds the 2500 ms timeout due to high LLM processing time; implement **streaming** (sending LLM tokens to TTS immediately as they are generated) to reduce perceived and t", "failure_class": null, "finish_reason": "stop", "latency_ms": 285.595178604126, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "text_checks": [ { "name": "identifies_budget_overrun", "passed": true }, { "name": "offers_latency_fix", "passed": true } ], "usage": { "completion_tokens": 52, "prompt_tokens": 78, "prompt_tokens_details": null, "total_tokens": 130 } }, { "attempt": 2, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "The total latency (3000 ms) exceeds the 2500 ms timeout due to the LLM's high processing time; implement **LLM streaming** to begin TTS as soon as the first tokens are generated.", "content_excerpt": "The total latency (3000 ms) exceeds the 2500 ms timeout due to the LLM's high processing time; implement **LLM streaming** to begin TTS as soon as the first tokens are generated.", "failure_class": "deterministic_check_failed", "finish_reason": "stop", "latency_ms": 263.05294036865234, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "failed", "text_checks": [ { "name": "identifies_budget_overrun", "passed": true }, { "name": "offers_latency_fix", "passed": false } ], "usage": { "completion_tokens": 49, "prompt_tokens": 78, "prompt_tokens_details": null, "total_tokens": 127 } }, { "attempt": 3, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "The total latency (3000 ms) exceeds the 2500 ms timeout due to the slow LLM response, which can be fixed by implementing LLM streaming to begin TTS generation before the full text is complete.", "content_excerpt": "The total latency (3000 ms) exceeds the 2500 ms timeout due to the slow LLM response, which can be fixed by implementing LLM streaming to begin TTS generation before the full text is complete.", "failure_class": "deterministic_check_failed", "finish_reason": "stop", "latency_ms": 261.11721992492676, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "failed", "text_checks": [ { "name": "identifies_budget_overrun", "passed": true }, { "name": "offers_latency_fix", "passed": false } ], "usage": { "completion_tokens": 48, "prompt_tokens": 78, "prompt_tokens_details": null, "total_tokens": 126 } } ], "error": "pass rate below threshold", "id": "parallel_timeout_triage", "pass_count": 1, "pass_rate": 0.3333333333333333, "required_pass_rate": 1.0, "status": "failed", "validator": "deterministic_text_checks" } ], "status": "failed" }, "run_id": "fast-bakeoff-20260716T231123Z", "schema": "anvil-serving.fast-tier-bakeoff/v1", "score_inputs": { "e2e_p50_ms": 1080.977439880371, "intelligence_pass_rate": 0.5, "operational_fit_notes": [ "endpoint was already loaded; benchmark did not start or stop serves" ], "session_recall_passed": true, "thinking_mode": "disabled", "tool_call_passed": true, "ttft_p50_ms": 744.2588806152344, "usable_context_tokens": 24000, "voice_latency_ms": null }, "selection": { "context_targets": [ 8192, 24000 ], "endpoint_already_loaded": true, "requests_per_context": 1, "suites": [ "chat", "context", "tool", "session", "intelligence" ] }, "session": { "checks": [ { "attempt_count": 3, "attempts": [ { "attempt": 1, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "RIVER-918", "content_excerpt": "RIVER-918", "expected": "RIVER-918", "failure_class": null, "finish_reason": "stop", "latency_ms": 90.07120132446289, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "usage": { "completion_tokens": 7, "prompt_tokens": 52, "prompt_tokens_details": null, "total_tokens": 59 } }, { "attempt": 2, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "RIVER-918", "content_excerpt": "RIVER-918", "expected": "RIVER-918", "failure_class": null, "finish_reason": "stop", "latency_ms": 66.98369979858398, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "usage": { "completion_tokens": 7, "prompt_tokens": 52, "prompt_tokens_details": null, "total_tokens": 59 } }, { "attempt": 3, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "RIVER-918", "content_excerpt": "RIVER-918", "expected": "RIVER-918", "failure_class": null, "finish_reason": "stop", "latency_ms": 75.42586326599121, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "usage": { "completion_tokens": 7, "prompt_tokens": 52, "prompt_tokens_details": null, "total_tokens": 59 } } ], "name": "single_request_multiturn_recall", "pass_count": 3, "pass_rate": 1.0, "required_pass_rate": 1.0, "status": "passed" } ], "status": "passed" }, "source_recipe": { "ref": null, "serve_command": null }, "suites": {}, "thinking": { "chat_template_kwargs": { "enable_thinking": false }, "control_evidence": "C:\\Users\\operator\\ai-code\\anvil-serving\\docs\\findings\\2026-07-16-gemma4-nvfp4-evidence\\nvfp4-thinking-control.json", "control_evidence_sha256": "76122ddb42cbb489bc023164a2693a741215b257d949c18d8663981903c2e99b", "control_mechanism": "chat_template_kwargs", "control_requested": { "enable_thinking": false }, "control_status": "supported", "mode": "disabled", "reasoning_effort": null, "unsupported": false }, "timing": { "chat": { "e2e_p50_ms": 1080.977439880371, "e2e_p95_ms": 1350.9016036987305, "output_tokens": 85, "ttft_p50_ms": 744.2588806152344, "ttft_p95_ms": 1220.9112644195557 }, "wall_ms": 4680.557727813721 }, "tool": { "checks": [ { "attempt_count": 3, "attempts": [ { "arguments": { "zip": "98101" }, "attempt": 1, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "", "content_excerpt": "", "failure_class": null, "finish_reason": "tool_calls", "latency_ms": 143.7819004058838, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "tool_call_count": 1, "usage": { "completion_tokens": 21, "prompt_tokens": 83, "prompt_tokens_details": null, "total_tokens": 104 }, "valid_tool_call_count": 1, "validation_errors": [] }, { "arguments": { "zip": "98101" }, "attempt": 2, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "", "content_excerpt": "", "failure_class": null, "finish_reason": "tool_calls", "latency_ms": 140.35701751708984, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "tool_call_count": 1, "usage": { "completion_tokens": 21, "prompt_tokens": 83, "prompt_tokens_details": null, "total_tokens": 104 }, "valid_tool_call_count": 1, "validation_errors": [] }, { "arguments": { "zip": "98101" }, "attempt": 3, "budget": { "legacy_max_tokens_as_visible": false, "max_completion_tokens": 512, "reasoning_headroom_tokens": 0, "visible_answer_tokens": 512 }, "content": "", "content_excerpt": "", "failure_class": null, "finish_reason": "tool_calls", "latency_ms": 142.6999568939209, "reasoning_chars": 0, "reasoning_excerpt": "", "reasoning_field": null, "reasoning_tokens": null, "status": "passed", "tool_call_count": 1, "usage": { "completion_tokens": 21, "prompt_tokens": 83, "prompt_tokens_details": null, "total_tokens": 104 }, "valid_tool_call_count": 1, "validation_errors": [] } ], "expected_arguments": { "zip": "98101" }, "expected_function": "record_weather_zip", "name": "openai_tool_call_smoke", "pass_count": 3, "pass_rate": 1.0, "required_pass_rate": 1.0, "status": "passed" } ], "status": "passed" }, "voice": { "llm_latency_ms": null, "status": "not_run", "stt_latency_ms": null, "total_turn_latency_ms": null, "tts_latency_ms": null } }