| { |
| "kind": "Select-v3-native-Kompress-v2-comparison-evidence-v1", |
| "user_date": "2026-11-04", |
| "general_superiority_established": false, |
| "qualification": false, |
| "historical_comparison": { |
| "status": "REUSED_HISTORICAL_DEV_OBSERVATIONS_NOT_QUALIFICATION", |
| "new_benchmark_runs": 0, |
| "model_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "hardware": "NVIDIA GeForce RTX 5080", |
| "kompress_scope_sha256": "8c13f887177adcf31fdfba03c6b5894598bfd6b33b74f086b757ec2eddb66332", |
| "v3_scope_sha256": "711c176b58492358dc0d4c7de4d79f7b1b20112d468ed8a5501d3f34d35ec3b3", |
| "v3_export_manifest_sha256": "19b10e408935e046946125af5f06994e0d034efc53ba5fb0dfdbe0b40f7053e1", |
| "kompress_export_manifest_sha256": "2774d37b3832201e59990182bc264fe15ca6383f3577264d94f9699e7816ddac", |
| "headroom_source_version": "source-1cf496612e781ef8d67ff87ee4f78037492f0bc7", |
| "parameter_counts": { |
| "v3": 149606401, |
| "kompress": 149999875 |
| }, |
| "precision": { |
| "live_parameter_dtype": "float32", |
| "CUDA_autocast": "bfloat16", |
| "cache_enabled": false, |
| "TF32": false, |
| "torch_threads": 1, |
| "GPU": 0, |
| "v3_attention": "admitted-SDPA", |
| "kompress_attention": "native-EAGER" |
| }, |
| "software": { |
| "huggingface_hub": "1.33.0", |
| "numpy": "2.5.3", |
| "peft": "0.21.2", |
| "safetensors": "0.8.0", |
| "tokenizers": "0.23.2", |
| "torch": "2.11.0+cu128", |
| "transformers": "5.19.0" |
| }, |
| "clock": "WARM only: start before fresh source JSON/span grouping, exact o200k count and .25/.5 target derivation; include request serialization/IPC, actual model tokenization/forward, API budget adaptation OR native allocation/recount, native child/relay/guardian wait, delivery and caller output recount. Fixed process/tokenizer/model startup+one case0 warmup disclosed separately, NOT cold cost. No result/source/score cache.", |
| "data": "One exposed DEV source per workload at two budgets; JSON/logs share source population; no independent complete-support truth or representative qualification set.", |
| "runtime": "Historical serialized research runners, not latency certification for the included raw-checkpoint quickstart. Kompress native-EAGER export was limited to batch1..8 and sequence5..512; native API guards on, CCR off; not CPU/INT8.", |
| "eligible_cells": 4, |
| "six_refusals_counted_as_wins": false, |
| "more_than_ten_percent_lower_recorded_times": 3, |
| "earlier_screens": { |
| "8c13f887": "0/4", |
| "a8948ec8": "0/4", |
| "ca18f5cd": "1/4", |
| "711c176b": "3/4" |
| }, |
| "not_contemporaneous_paired_causal_or_statistical": true, |
| "independent_semantic_health_established": false, |
| "quality_p50_p95_cold_or90percent_claim": false, |
| "qualified_superiority": false, |
| "source_receipts": [ |
| { |
| "path": "tmp/select-v3-native-guardian-cost-audit-v1/result.json", |
| "sha256": "3ee330fa203951ff2e9be59e05230d9c033bde253da11c8c80284d5acf147a7c" |
| }, |
| { |
| "path": "tmp/select-v3-native-guardian-cost-v1/result.json", |
| "sha256": "402dec4660ce4c756c76af1af5b09f65700c8f193957a82c605ea60ab8a55e95" |
| }, |
| { |
| "path": "tmp/select-v3-native-guardian-cost-v1/scope.json", |
| "sha256": "711c176b58492358dc0d4c7de4d79f7b1b20112d468ed8a5501d3f34d35ec3b3" |
| }, |
| { |
| "path": "tmp/select-v3-native-guardian-cost-v1/cases.json", |
| "sha256": "abc70207fdf4774c2a9d0f8ad6d9090c3dad5a51ab7896ca827f6081aa53ba37" |
| }, |
| { |
| "path": "tmp/select-v3-kompress-resident-cost-audit-v1/result.json", |
| "sha256": "dd279335902db0e1cf57af20781b95d71b19f0ca84864cbd71dda6411aec45d4" |
| }, |
| { |
| "path": "tmp/select-v3-kompress-resident-cost-v1/result.json", |
| "sha256": "ffbec64a9fb7afc057e5bf97667ff376845c2c9c1f542e14385937c670d47745" |
| }, |
| { |
| "path": "tmp/select-v3-kompress-resident-cost-v1/scope.json", |
| "sha256": "8c13f887177adcf31fdfba03c6b5894598bfd6b33b74f086b757ec2eddb66332" |
| }, |
| { |
| "path": "tmp/select-v3-kompress-resident-cost-v1/cases.json", |
| "sha256": "abc70207fdf4774c2a9d0f8ad6d9090c3dad5a51ab7896ca827f6081aa53ba37" |
| }, |
| { |
| "path": "tmp/select-v3-kompress-budget-review-v1/result.json", |
| "sha256": "5ff4b373715ed3891f6cfdd7905ebee67f1b9d137ac84b5311dc5c4b93aa17fb" |
| } |
| ], |
| "rows": [ |
| { |
| "case_index": 0, |
| "workload": "code", |
| "retention_budget": "0.25", |
| "input_tokens": 4411, |
| "target_tokens": 1102, |
| "v3_output_tokens": 1079, |
| "v3_tokens_removed": 3332, |
| "v3_request_seconds": "0.1589187", |
| "kompress_recorded_seconds": "0.2036338", |
| "kompress_budget_eligible": false, |
| "kompress_recorded_seconds_are_refusal_cost": true, |
| "kompress_returned_tokens": 4411, |
| "kompress_native_attempt_tokens": [ |
| 2032, |
| 1747, |
| 1642, |
| 1607 |
| ], |
| "v3_vs_kompress_time_change_percent": null, |
| "comparison": "Inconclusive", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 75.53842666062117 |
| }, |
| { |
| "case_index": 1, |
| "workload": "code", |
| "retention_budget": "0.5", |
| "input_tokens": 4411, |
| "target_tokens": 2205, |
| "v3_output_tokens": 2205, |
| "v3_tokens_removed": 2206, |
| "v3_request_seconds": "0.1460882", |
| "kompress_recorded_seconds": "0.2003043", |
| "kompress_budget_eligible": true, |
| "kompress_recorded_seconds_are_refusal_cost": false, |
| "kompress_returned_tokens": 2149, |
| "kompress_native_attempt_tokens": [ |
| 2469, |
| 2301, |
| 2215, |
| 2149 |
| ], |
| "v3_vs_kompress_time_change_percent": "-27.06686776070209176737593751", |
| "comparison": "27.1% lower time", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 50.01133529811834 |
| }, |
| { |
| "case_index": 2, |
| "workload": "json", |
| "retention_budget": "0.25", |
| "input_tokens": 1303, |
| "target_tokens": 325, |
| "v3_output_tokens": 308, |
| "v3_tokens_removed": 995, |
| "v3_request_seconds": "0.0628606", |
| "kompress_recorded_seconds": "0.0804463", |
| "kompress_budget_eligible": false, |
| "kompress_recorded_seconds_are_refusal_cost": true, |
| "kompress_returned_tokens": 1303, |
| "kompress_native_attempt_tokens": [ |
| 984, |
| 984, |
| 984, |
| 984 |
| ], |
| "v3_vs_kompress_time_change_percent": null, |
| "comparison": "Inconclusive", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 76.36224098234842 |
| }, |
| { |
| "case_index": 3, |
| "workload": "json", |
| "retention_budget": "0.5", |
| "input_tokens": 1303, |
| "target_tokens": 651, |
| "v3_output_tokens": 651, |
| "v3_tokens_removed": 652, |
| "v3_request_seconds": "0.0658957", |
| "kompress_recorded_seconds": "0.0787157", |
| "kompress_budget_eligible": false, |
| "kompress_recorded_seconds_are_refusal_cost": true, |
| "kompress_returned_tokens": 1303, |
| "kompress_native_attempt_tokens": [ |
| 1020, |
| 984, |
| 984, |
| 984 |
| ], |
| "v3_vs_kompress_time_change_percent": null, |
| "comparison": "Inconclusive", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 50.03837298541826 |
| }, |
| { |
| "case_index": 4, |
| "workload": "logs", |
| "retention_budget": "0.25", |
| "input_tokens": 1103, |
| "target_tokens": 275, |
| "v3_output_tokens": 275, |
| "v3_tokens_removed": 828, |
| "v3_request_seconds": "0.0589192", |
| "kompress_recorded_seconds": "0.07833", |
| "kompress_budget_eligible": false, |
| "kompress_recorded_seconds_are_refusal_cost": true, |
| "kompress_returned_tokens": 1103, |
| "kompress_native_attempt_tokens": [ |
| 895, |
| 895, |
| 895, |
| 895 |
| ], |
| "v3_vs_kompress_time_change_percent": null, |
| "comparison": "Inconclusive", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 75.06799637352674 |
| }, |
| { |
| "case_index": 5, |
| "workload": "logs", |
| "retention_budget": "0.5", |
| "input_tokens": 1103, |
| "target_tokens": 551, |
| "v3_output_tokens": 503, |
| "v3_tokens_removed": 600, |
| "v3_request_seconds": "0.0579552", |
| "kompress_recorded_seconds": "0.0785184", |
| "kompress_budget_eligible": false, |
| "kompress_recorded_seconds_are_refusal_cost": true, |
| "kompress_returned_tokens": 1103, |
| "kompress_native_attempt_tokens": [ |
| 913, |
| 895, |
| 895, |
| 895 |
| ], |
| "v3_vs_kompress_time_change_percent": null, |
| "comparison": "Inconclusive", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 54.39709882139619 |
| }, |
| { |
| "case_index": 6, |
| "workload": "multi-turn", |
| "retention_budget": "0.25", |
| "input_tokens": 587, |
| "target_tokens": 146, |
| "v3_output_tokens": 39, |
| "v3_tokens_removed": 548, |
| "v3_request_seconds": "0.069653", |
| "kompress_recorded_seconds": "0.0906459", |
| "kompress_budget_eligible": true, |
| "kompress_recorded_seconds_are_refusal_cost": false, |
| "kompress_returned_tokens": 144, |
| "kompress_native_attempt_tokens": [ |
| 255, |
| 176, |
| 153, |
| 144 |
| ], |
| "v3_vs_kompress_time_change_percent": "-23.15923831083369462932134824", |
| "comparison": "23.2% lower time", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 93.35604770017036 |
| }, |
| { |
| "case_index": 7, |
| "workload": "multi-turn", |
| "retention_budget": "0.5", |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "v3_output_tokens": 235, |
| "v3_tokens_removed": 352, |
| "v3_request_seconds": "0.0693616", |
| "kompress_recorded_seconds": "0.0644106", |
| "kompress_budget_eligible": true, |
| "kompress_recorded_seconds_are_refusal_cost": false, |
| "kompress_returned_tokens": 284, |
| "kompress_native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ], |
| "v3_vs_kompress_time_change_percent": "7.686623009256240432475400000", |
| "comparison": "7.7% higher time", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 59.96592844974446 |
| }, |
| { |
| "case_index": 8, |
| "workload": "retrieval", |
| "retention_budget": "0.25", |
| "input_tokens": 4498, |
| "target_tokens": 1124, |
| "v3_output_tokens": 1115, |
| "v3_tokens_removed": 3383, |
| "v3_request_seconds": "0.1361903", |
| "kompress_recorded_seconds": "0.2695771", |
| "kompress_budget_eligible": false, |
| "kompress_recorded_seconds_are_refusal_cost": true, |
| "kompress_returned_tokens": 4498, |
| "kompress_native_attempt_tokens": [ |
| 1950, |
| 1578, |
| 1428, |
| 1347 |
| ], |
| "v3_vs_kompress_time_change_percent": null, |
| "comparison": "Inconclusive", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 75.21120497999111 |
| }, |
| { |
| "case_index": 9, |
| "workload": "retrieval", |
| "retention_budget": "0.5", |
| "input_tokens": 4498, |
| "target_tokens": 2249, |
| "v3_output_tokens": 2217, |
| "v3_tokens_removed": 2281, |
| "v3_request_seconds": "0.139291", |
| "kompress_recorded_seconds": "0.201564", |
| "kompress_budget_eligible": true, |
| "kompress_recorded_seconds_are_refusal_cost": false, |
| "kompress_returned_tokens": 2219, |
| "kompress_native_attempt_tokens": [ |
| 2651, |
| 2361, |
| 2219 |
| ], |
| "v3_vs_kompress_time_change_percent": "-30.89490186739695580560020639", |
| "comparison": "30.9% lower time", |
| "semantic_health_or_quality_established": false, |
| "v3_tokens_saved_percent": 50.711427301022674 |
| } |
| ] |
| }, |
| "historical_50percent_summary": { |
| "eligible_rows": 3, |
| "V3_lower_recorded_time_rows": [ |
| "code", |
| "retrieval" |
| ], |
| "V3_higher_recorded_time_rows": [ |
| "multi-turn" |
| ], |
| "comparator_budget_refusal_rows": [ |
| "json", |
| "logs" |
| ], |
| "causal_paired_trial": false, |
| "quality_compared": false |
| }, |
| "publisher_data_review": { |
| "status": "PUBLIC_SHARED_PILOT_FOUND_NOT_LABELED_F1_TEST", |
| "model_repository": "chopratejas/kompress-v2-base", |
| "model_card_revision": "e00cc5e60ff73b01674ef8fa6da6c5acf6d54181", |
| "public_published_metrics": { |
| "threshold": 0.5, |
| "token_precision": 0.94, |
| "token_recall": 0.901, |
| "token_f1": 0.92, |
| "must_keep_recall": 0.989, |
| "keep_rate": 0.866, |
| "test_rows": 12697, |
| "scored_subwords": 16827490, |
| "dataset_name": "dataset_v2_v2.1-f00", |
| "hardware": "H100", |
| "compute": "PyTorch BF16", |
| "source": "https://huggingface.co/chopratejas/kompress-v2-base/blob/e00cc5e60ff73b01674ef8fa6da6c5acf6d54181/README.md", |
| "third_party_reported_not_reproduced": true |
| }, |
| "installed_kompress_merged_sha256": "04acb33493b84767f93767784249d2cbdf01e30bdde353b94105984f0617c30b", |
| "installed_weights_match_public_model": true, |
| "exact_labeled_test_split_publicly_located": false, |
| "checked_locations": [ |
| "publisher model file inventory", |
| "publisher-owned HF dataset listing", |
| "public Headroom recursive tree", |
| "Kompress training GitHub repository: unauthenticated404" |
| ], |
| "absence_limit": "No public copy found in these checked locations; not a proof the data cannot be obtained.", |
| "shared_pilot": { |
| "repository": "chopratejas/headroom-datasets", |
| "revision": "2fd099f9f565f1af59e22a6a51671faff6a227f9", |
| "rows": 30, |
| "split": "train", |
| "license": "CC-BY-4.0", |
| "source": "synthetic", |
| "downloaded_bytes": 5950123, |
| "sha256": "ef485e1918b0df66976238e72adb3300196378de244caf9032547a545e6f65f8", |
| "permission": "Explicit USER evaluation-only exception, no training.", |
| "attribution": "Headroom Pilot by chopratejas; https://huggingface.co/datasets/chopratejas/headroom-datasets; CC-BY-4.0 https://creativecommons.org/licenses/by/4.0/" |
| }, |
| "adaptation": "5 first-id category cases; flatten three recorded tool outputs in order. Not the full Anthropic/proxy/cache benchmark, not token keep/drop labels.", |
| "new_f1_superiority_claim": false, |
| "prior_failed_attempt": "pilot-terminal.json", |
| "corrected_attempt": "tmp/select-v3-kompress-shared-pilot-v2; awaiting explicit fresh approval" |
| }, |
| "shared_pilot": { |
| "status": "PARTIAL_V3_COMPLETE_COMPARATOR_SETUP_FAILED_NO_QA", |
| "dataset": { |
| "repository": "chopratejas/headroom-datasets", |
| "revision": "2fd099f9f565f1af59e22a6a51671faff6a227f9", |
| "rows": 30, |
| "split": "train", |
| "license": "CC-BY-4.0", |
| "source": "synthetic", |
| "downloaded_bytes": 5950123, |
| "sha256": "ef485e1918b0df66976238e72adb3300196378de244caf9032547a545e6f65f8", |
| "permission": "Explicit USER evaluation-only exception, no training.", |
| "attribution": "Headroom Pilot by chopratejas; https://huggingface.co/datasets/chopratejas/headroom-datasets; CC-BY-4.0 https://creativecommons.org/licenses/by/4.0/" |
| }, |
| "selection": { |
| "dataset": "chopratejas/headroom-datasets", |
| "revision": "2fd099f9f565f1af59e22a6a51671faff6a227f9", |
| "license": "CC-BY-4.0", |
| "permission": "explicit user exception for evaluation only; no training", |
| "selection": "lexicographically smallest id in each of the five categories, frozen before model outcomes", |
| "ids": [ |
| "agentic_loop_001", |
| "code_agent_001", |
| "json_data_001", |
| "logs_001", |
| "rag_search_001" |
| ], |
| "dataset_sha256": "ef485e1918b0df66976238e72adb3300196378de244caf9032547a545e6f65f8", |
| "qualification": false |
| }, |
| "scope_sha256": "835ecd96375ba53d5f88ff8d4d3818815e20d3d2d34942eae3fe288163608594", |
| "approval_sha256": "b9b0907856a05b39834f62e6179890a38e6c9b444debb462d7be6cb3cdd049fd", |
| "scope_path": "tmp\\select-v3-kompress-shared-pilot-v2", |
| "compression_cells": [ |
| { |
| "id": "agentic_loop_001", |
| "arm": "v3", |
| "input_tokens": 41097, |
| "target_tokens": 20548, |
| "output_tokens": 20433, |
| "valid_budget": true, |
| "status": "compressed", |
| "request_seconds": 1.3425146999998105, |
| "payload_sha256": "cd26abe9a678bbac2380e92d2937ae9c3faf7b621ac84172aef722c406a8407b", |
| "receipt": { |
| "schema_version": "1.0", |
| "kind": "context_selection", |
| "selected": true, |
| "used_scorer": false, |
| "fallback_reason": "scorer unavailable", |
| "baseline_tokens": 41097, |
| "output_tokens": 20433, |
| "target_tokens": 20548, |
| "budget_met": true, |
| "kept_group_count": 115, |
| "estimator": { |
| "backend": "tiktoken", |
| "model": "o200k_base", |
| "is_exact": true |
| }, |
| "scorer_revision": null, |
| "kept_group_ids": [ |
| "block-000320", |
| "block-000400", |
| "block-000496", |
| "block-000576", |
| "block-000592", |
| "block-000624", |
| "block-000672", |
| "block-000800", |
| "block-000976", |
| "block-000992", |
| "block-001024", |
| "block-001040", |
| "block-001056", |
| "block-001072", |
| "block-001088", |
| "block-001104", |
| "block-001120", |
| "block-001216", |
| "block-001232", |
| "block-001280", |
| "block-001360", |
| "block-001376", |
| "block-001584", |
| "block-001632", |
| "block-001648", |
| "block-001696", |
| "block-001712", |
| "block-001728", |
| "block-001776", |
| "block-001792", |
| "block-001808", |
| "block-001952", |
| "block-001968", |
| "block-001984", |
| "block-002000", |
| "block-002016", |
| "block-002032", |
| "block-002048", |
| "block-002176", |
| "block-002272", |
| "block-002288", |
| "block-002368", |
| "block-002384", |
| "block-002416", |
| "block-002432", |
| "block-002448", |
| "block-002464", |
| "block-002480", |
| "block-002528", |
| "block-002560", |
| "block-002592", |
| "block-002656", |
| "block-002672", |
| "block-002688", |
| "block-002704", |
| "block-002736", |
| "block-002800", |
| "block-002832", |
| "block-002848", |
| "block-002864", |
| "block-002880", |
| "block-002896", |
| "block-002912", |
| "block-002928", |
| "block-002944", |
| "block-002960", |
| "block-002976", |
| "block-002992", |
| "block-003008", |
| "block-003024", |
| "block-003040", |
| "block-003056", |
| "block-003072", |
| "block-003088", |
| "block-003104", |
| "block-003120", |
| "block-003136", |
| "block-003152", |
| "block-003168", |
| "block-003184", |
| "block-003200", |
| "block-003216", |
| "block-003232", |
| "block-003248", |
| "block-003264", |
| "block-003280", |
| "block-003296", |
| "block-003312", |
| "block-003328", |
| "block-003344", |
| "block-003360", |
| "block-003376", |
| "block-003392", |
| "block-003408", |
| "block-003424", |
| "block-003440", |
| "block-003456", |
| "block-003472", |
| "block-003488", |
| "block-003504", |
| "block-003520", |
| "block-003536", |
| "block-003552", |
| "block-003568", |
| "block-003728", |
| "block-003744", |
| "block-003776", |
| "block-003808", |
| "block-003824", |
| "block-003840", |
| "block-003856", |
| "block-003872", |
| "block-003888", |
| "block-003904", |
| "block-003920" |
| ] |
| }, |
| "qualification": false, |
| "category": "agentic_loop", |
| "evidence_path": "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\0.json", |
| "evidence_sha256": "b1560d960f7b508a590b5e6c1d3567536cc7004e1a8a55a8433c3e6bc9429593", |
| "requested_literal_presence": { |
| "id": "agentic_loop_001", |
| "input_tokens": 41097, |
| "output_tokens": 20433, |
| "actual_reduction_percent": 50.281042411854884, |
| "requested_values": 3, |
| "values_present_in_raw": 3, |
| "values_present_in_selected": 3, |
| "all_values_present": true |
| } |
| }, |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "kompress", |
| "input_tokens": 41097, |
| "target_tokens": 20548, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_comparator_setup_failed", |
| "scores": null, |
| "qualification": false |
| }, |
| { |
| "id": "code_agent_001", |
| "arm": "v3", |
| "input_tokens": 7141, |
| "target_tokens": 3570, |
| "output_tokens": 3568, |
| "valid_budget": true, |
| "status": "compressed", |
| "request_seconds": 0.31693759999870963, |
| "payload_sha256": "53d7dbfe70b31f4d2464e60ab104e00a5eb4ecfd462ce04489ec81b58535c4cc", |
| "receipt": { |
| "schema_version": "1.0", |
| "kind": "context_selection", |
| "selected": true, |
| "used_scorer": false, |
| "fallback_reason": "scorer unavailable", |
| "baseline_tokens": 7141, |
| "output_tokens": 3568, |
| "target_tokens": 3570, |
| "budget_met": true, |
| "kept_group_count": 22, |
| "estimator": { |
| "backend": "tiktoken", |
| "model": "o200k_base", |
| "is_exact": true |
| }, |
| "scorer_revision": null, |
| "kept_group_ids": [ |
| "block-000000", |
| "block-000016", |
| "block-000096", |
| "block-000128", |
| "block-000144", |
| "block-000176", |
| "block-000208", |
| "block-000224", |
| "block-000256", |
| "block-000272", |
| "block-000288", |
| "block-000304", |
| "block-000336", |
| "block-000368", |
| "block-000384", |
| "block-000416", |
| "block-000464", |
| "block-000576", |
| "block-000608", |
| "block-000624", |
| "block-000672", |
| "block-000688" |
| ] |
| }, |
| "qualification": false, |
| "category": "code_agent", |
| "evidence_path": "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\1.json", |
| "evidence_sha256": "31bcdb68db4007600e82c413df971cca6e52ee21334f33a4e3234d098f8b9265", |
| "requested_literal_presence": { |
| "id": "code_agent_001", |
| "input_tokens": 7141, |
| "output_tokens": 3568, |
| "actual_reduction_percent": 50.03500910236662, |
| "requested_values": 2, |
| "values_present_in_raw": 2, |
| "values_present_in_selected": 2, |
| "all_values_present": true |
| } |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "kompress", |
| "input_tokens": 7141, |
| "target_tokens": 3570, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_comparator_setup_failed", |
| "scores": null, |
| "qualification": false |
| }, |
| { |
| "id": "json_data_001", |
| "arm": "v3", |
| "input_tokens": 11345, |
| "target_tokens": 5672, |
| "output_tokens": 5637, |
| "valid_budget": true, |
| "status": "compressed", |
| "request_seconds": 0.36647290000109933, |
| "payload_sha256": "dba4fcf0667fd35da788c6018cc71570943f24915f175cd5288822160f4c36a1", |
| "receipt": { |
| "schema_version": "1.0", |
| "kind": "context_selection", |
| "selected": true, |
| "used_scorer": false, |
| "fallback_reason": "scorer unavailable", |
| "baseline_tokens": 11345, |
| "output_tokens": 5637, |
| "target_tokens": 5672, |
| "budget_met": true, |
| "kept_group_count": 44, |
| "estimator": { |
| "backend": "tiktoken", |
| "model": "o200k_base", |
| "is_exact": true |
| }, |
| "scorer_revision": null, |
| "kept_group_ids": [ |
| "block-000000", |
| "block-000016", |
| "block-000032", |
| "block-000048", |
| "block-000080", |
| "block-000096", |
| "block-000112", |
| "block-000128", |
| "block-000144", |
| "block-000160", |
| "block-000176", |
| "block-000192", |
| "block-000208", |
| "block-000224", |
| "block-000240", |
| "block-000256", |
| "block-000272", |
| "block-000288", |
| "block-000304", |
| "block-000320", |
| "block-000352", |
| "block-000368", |
| "block-000384", |
| "block-000464", |
| "block-000496", |
| "block-000512", |
| "block-000528", |
| "block-000656", |
| "block-000672", |
| "block-000752", |
| "block-000768", |
| "block-000800", |
| "block-000816", |
| "block-000912", |
| "block-000928", |
| "block-000960", |
| "block-000992", |
| "block-001024", |
| "block-001040", |
| "block-001056", |
| "block-001072", |
| "block-001168", |
| "block-001312", |
| "block-001360" |
| ] |
| }, |
| "qualification": false, |
| "category": "json_data", |
| "evidence_path": "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\2.json", |
| "evidence_sha256": "beef722fe47fc637062a0adacb62f3df19a18c811bccf345eb5d0b36c2c946fc", |
| "requested_literal_presence": { |
| "id": "json_data_001", |
| "input_tokens": 11345, |
| "output_tokens": 5637, |
| "actual_reduction_percent": 50.312913177611286, |
| "requested_values": 3, |
| "values_present_in_raw": 3, |
| "values_present_in_selected": 3, |
| "all_values_present": true |
| } |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "kompress", |
| "input_tokens": 11345, |
| "target_tokens": 5672, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_comparator_setup_failed", |
| "scores": null, |
| "qualification": false |
| }, |
| { |
| "id": "logs_001", |
| "arm": "v3", |
| "input_tokens": 39682, |
| "target_tokens": 19841, |
| "output_tokens": 19751, |
| "valid_budget": true, |
| "status": "compressed", |
| "request_seconds": 1.0018507999993744, |
| "payload_sha256": "bfcba29be42f721eae7f2943b604dafaac8f7c4b7e2d1940387e1820ff2e3fe8", |
| "receipt": { |
| "schema_version": "1.0", |
| "kind": "context_selection", |
| "selected": true, |
| "used_scorer": false, |
| "fallback_reason": "scorer unavailable", |
| "baseline_tokens": 39682, |
| "output_tokens": 19751, |
| "target_tokens": 19841, |
| "budget_met": true, |
| "kept_group_count": 124, |
| "estimator": { |
| "backend": "tiktoken", |
| "model": "o200k_base", |
| "is_exact": true |
| }, |
| "scorer_revision": null, |
| "kept_group_ids": [ |
| "block-000128", |
| "block-000144", |
| "block-000160", |
| "block-000176", |
| "block-000192", |
| "block-000208", |
| "block-000224", |
| "block-000240", |
| "block-000304", |
| "block-000320", |
| "block-000336", |
| "block-000352", |
| "block-000368", |
| "block-000384", |
| "block-000416", |
| "block-000464", |
| "block-000496", |
| "block-000544", |
| "block-000560", |
| "block-000592", |
| "block-000640", |
| "block-000784", |
| "block-000864", |
| "block-000912", |
| "block-000928", |
| "block-000960", |
| "block-000976", |
| "block-000992", |
| "block-001008", |
| "block-001024", |
| "block-001056", |
| "block-001072", |
| "block-001088", |
| "block-001104", |
| "block-001248", |
| "block-001280", |
| "block-001296", |
| "block-001344", |
| "block-001376", |
| "block-001488", |
| "block-001520", |
| "block-001536", |
| "block-001552", |
| "block-001568", |
| "block-001584", |
| "block-001600", |
| "block-001616", |
| "block-001632", |
| "block-001648", |
| "block-001680", |
| "block-001696", |
| "block-001712", |
| "block-001728", |
| "block-001744", |
| "block-001760", |
| "block-001776", |
| "block-001792", |
| "block-001808", |
| "block-001824", |
| "block-001840", |
| "block-001856", |
| "block-001872", |
| "block-001888", |
| "block-001904", |
| "block-001920", |
| "block-001936", |
| "block-001968", |
| "block-001984", |
| "block-002000", |
| "block-002016", |
| "block-002112", |
| "block-002160", |
| "block-002208", |
| "block-002256", |
| "block-002320", |
| "block-002368", |
| "block-002384", |
| "block-002416", |
| "block-002448", |
| "block-002464", |
| "block-002480", |
| "block-002512", |
| "block-002656", |
| "block-002736", |
| "block-002928", |
| "block-002944", |
| "block-002960", |
| "block-002976", |
| "block-002992", |
| "block-003008", |
| "block-003040", |
| "block-003072", |
| "block-003088", |
| "block-003120", |
| "block-003136", |
| "block-003152", |
| "block-003168", |
| "block-003200", |
| "block-003344", |
| "block-003376", |
| "block-003392", |
| "block-003408", |
| "block-003520", |
| "block-003552", |
| "block-003568", |
| "block-003632", |
| "block-003664", |
| "block-003680", |
| "block-003696", |
| "block-003712", |
| "block-003728", |
| "block-003744", |
| "block-003760", |
| "block-003776", |
| "block-003792", |
| "block-003808", |
| "block-003824", |
| "block-003840", |
| "block-003856", |
| "block-003872", |
| "block-003888", |
| "block-003920", |
| "block-003968", |
| "block-003984" |
| ] |
| }, |
| "qualification": false, |
| "category": "logs", |
| "evidence_path": "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\3.json", |
| "evidence_sha256": "d4ba37598067b3daf0d8075002584fb809bf2ccadab05d5ab32bf6a86dc26307", |
| "requested_literal_presence": { |
| "id": "logs_001", |
| "input_tokens": 39682, |
| "output_tokens": 19751, |
| "actual_reduction_percent": 50.22680308452194, |
| "requested_values": 2, |
| "values_present_in_raw": 2, |
| "values_present_in_selected": 2, |
| "all_values_present": true |
| } |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "kompress", |
| "input_tokens": 39682, |
| "target_tokens": 19841, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_comparator_setup_failed", |
| "scores": null, |
| "qualification": false |
| }, |
| { |
| "id": "rag_search_001", |
| "arm": "v3", |
| "input_tokens": 6435, |
| "target_tokens": 3217, |
| "output_tokens": 2908, |
| "valid_budget": true, |
| "status": "compressed", |
| "request_seconds": 0.24580069999865373, |
| "payload_sha256": "7be46d38ac8b509c668aa21eb5cafa92ee799351951650fd1323d843b7b49081", |
| "receipt": { |
| "schema_version": "1.0", |
| "kind": "context_selection", |
| "selected": true, |
| "used_scorer": false, |
| "fallback_reason": "scorer unavailable", |
| "baseline_tokens": 6435, |
| "output_tokens": 2908, |
| "target_tokens": 3217, |
| "budget_met": true, |
| "kept_group_count": 4, |
| "estimator": { |
| "backend": "tiktoken", |
| "model": "o200k_base", |
| "is_exact": true |
| }, |
| "scorer_revision": null, |
| "kept_group_ids": [ |
| "block-000032", |
| "block-000048", |
| "block-000064", |
| "block-000112" |
| ] |
| }, |
| "qualification": false, |
| "category": "rag_search", |
| "evidence_path": "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\4.json", |
| "evidence_sha256": "fea19abd0746ccb763835fa824cbf8e3de896da5dc97a55d12f21a6a9daa6922", |
| "requested_literal_presence": { |
| "id": "rag_search_001", |
| "input_tokens": 6435, |
| "output_tokens": 2908, |
| "actual_reduction_percent": 54.80963480963481, |
| "requested_values": 2, |
| "values_present_in_raw": 2, |
| "values_present_in_selected": 2, |
| "all_values_present": true |
| } |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "kompress", |
| "input_tokens": 6435, |
| "target_tokens": 3217, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_comparator_setup_failed", |
| "scores": null, |
| "qualification": false |
| } |
| ], |
| "answer_cells": [ |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "raw", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "v3", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "kompress", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "raw", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "v3", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "kompress", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "raw", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "v3", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "kompress", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "raw", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "v3", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "kompress", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "raw", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "v3", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "kompress", |
| "status": "unexecuted_after_comparator_setup_failure", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| } |
| ], |
| "overall_P_R_F1": null, |
| "completed_V3_compression_cases": 5, |
| "completed_V3_scoring_windows": 17, |
| "Kompress_model_loads": 0, |
| "Kompress_compression_calls": 0, |
| "answerer_requests": 0, |
| "requested_literal_substring_presence": { |
| "kind": "publisher-requested-literal-value-presence-after-fixed-selection", |
| "is_answer_accuracy_or_f1": false, |
| "rows": [ |
| { |
| "id": "agentic_loop_001", |
| "input_tokens": 41097, |
| "output_tokens": 20433, |
| "actual_reduction_percent": 50.281042411854884, |
| "requested_values": 3, |
| "values_present_in_raw": 3, |
| "values_present_in_selected": 3, |
| "all_values_present": true |
| }, |
| { |
| "id": "code_agent_001", |
| "input_tokens": 7141, |
| "output_tokens": 3568, |
| "actual_reduction_percent": 50.03500910236662, |
| "requested_values": 2, |
| "values_present_in_raw": 2, |
| "values_present_in_selected": 2, |
| "all_values_present": true |
| }, |
| { |
| "id": "json_data_001", |
| "input_tokens": 11345, |
| "output_tokens": 5637, |
| "actual_reduction_percent": 50.312913177611286, |
| "requested_values": 3, |
| "values_present_in_raw": 3, |
| "values_present_in_selected": 3, |
| "all_values_present": true |
| }, |
| { |
| "id": "logs_001", |
| "input_tokens": 39682, |
| "output_tokens": 19751, |
| "actual_reduction_percent": 50.22680308452194, |
| "requested_values": 2, |
| "values_present_in_raw": 2, |
| "values_present_in_selected": 2, |
| "all_values_present": true |
| }, |
| { |
| "id": "rag_search_001", |
| "input_tokens": 6435, |
| "output_tokens": 2908, |
| "actual_reduction_percent": 54.80963480963481, |
| "requested_values": 2, |
| "values_present_in_raw": 2, |
| "values_present_in_selected": 2, |
| "all_values_present": true |
| } |
| ], |
| "all_value_cases": 5, |
| "cases": 5, |
| "qualification": false |
| }, |
| "literal_presence_is_not_answer_accuracy_or_f1": true, |
| "error": { |
| "status": "TERMINAL_NO_RETRY", |
| "error_type": "AttributeError", |
| "error": "'KompressConfig' object has no attribute 'batch_size'" |
| }, |
| "lifecycles": { |
| "v3": { |
| "worker_original_waited": true, |
| "guardian_original_waited": true, |
| "worker_exit": 0, |
| "guardian_exit": 0, |
| "stop_reason": null, |
| "action_completed": false, |
| "action_error_type": null, |
| "qualification": false |
| }, |
| "kompress": { |
| "worker_original_waited": true, |
| "guardian_original_waited": true, |
| "worker_exit": 1, |
| "guardian_exit": 0, |
| "stop_reason": null, |
| "action_completed": false, |
| "action_error_type": null, |
| "qualification": false |
| } |
| }, |
| "source_only_diagnosis": { |
| "status": "SOURCE_ONLY_NO_NEW_MODEL_CALLS", |
| "root_cause": "Current pilot copied an older failed smoke helper, which incorrectly checked model.config.batch_size; batch_size is a compress_batch parameter, not a config field.", |
| "native_config_fields": [ |
| "device", |
| "enable_ccr", |
| "model_id", |
| "chunk_words", |
| "score_threshold", |
| "min_input_words" |
| ], |
| "native_compress_batch_default": 32, |
| "existing_correct_helper": "E:\\Github\\tokenfold\\tmp\\select-v3-practical-release-residual-v3\\smoke_worker.py", |
| "existing_helper_sha256": "5dabb23f4145aec6c6dca3eb5840326292ed11e4203f473bd84a991a79a2c34e", |
| "passed_native_control_receipt": "E:\\Github\\tokenfold\\tmp\\select-v3-practical-release-residual-v3\\execution-terminal.json", |
| "passed_receipt_sha256": "75cfcaf41c2215a8d6e690945ed875f8cee0dc4e0f0a53e221ce9860d033197b", |
| "evidence_limit": "Existing functional smoke does not establish performance on the new pilot. New pilot is terminal; no implicit retry.", |
| "next_valid_measurement": "Fresh separately approved Kompress-plus-answerer-only scope may reuse the five completed immutable V3 outputs; never rescore V3 just to complete the comparator." |
| }, |
| "input_integrity": { |
| "status": "FROZEN_BENCHMARK_INPUTS_PRESERVED", |
| "requested_outcome_driven_input_refinement_performed": false, |
| "manufactured_superiority_claim": false, |
| "reason": "Results and inputs must remain auditable. Genuine advantages may be highlighted; any later curated illustration must be labeled separately, not treated as representative benchmark superiority.", |
| "scope_sha256": "835ecd96375ba53d5f88ff8d4d3818815e20d3d2d34942eae3fe288163608594", |
| "new_model_execution_after_terminal": false |
| }, |
| "clock_limit": "V3 recorded request_seconds exclude process/model load and frozen caller grouping; not the historical complete-request latency clock and not a paired comparison.", |
| "geometry_limit": "Four of five cases use expanded group/window geometry; no general quality claim.", |
| "next_execution": "No further model work approved. A new comparator-plus-answerer scope could reuse these V3 outputs without rescoring them.", |
| "qualification": false, |
| "training": 0, |
| "model_downloads": 0, |
| "new_runs_after_terminal": 0 |
| }, |
| "metric_warning": "V3 met 50% context caps on5 fixed pilot cases and all12 requested literal substrings remain somewhere in selected context. Literal presence is not correct answers, semantic binding, P/R/F1 or proof of superiority. Kompress setup failed in our copied stale harness before model load; not a comparator loss. Do not compare older V3 lookup answer F1=1.00 to published token-label Kompress F1=.920.", |
| "attribution": "Kompress v2 model card by chopratejas. Headroom Pilot by chopratejas, CC-BY-4.0; https://huggingface.co/datasets/chopratejas/headroom-datasets; https://creativecommons.org/licenses/by/4.0/. Raw pilot data is not bundled.", |
| "immutable_evidence_sha256": { |
| "tmp\\select-v3-kompress-readme-comparison-v1\\public-api-receipts.json": "1a4cde08fbfbc142ed3337e96132cb67169a29a6e0402e45f328b96aa98a1fa8", |
| "tmp\\select-v3-kompress-readme-comparison-v1\\source-review.json": "a8be0a7d4d60efaf1209821884976f263da1cd9baf92692c08ba27e5fcae42b5", |
| "tmp\\select-v3-kompress-readme-comparison-v1\\pilot-terminal.json": "f2166a491b2d028e17d4cf6bb79b5b97228c1f01954daf10b06981ca91218a43", |
| "tmp\\select-v3-kompress-readme-comparison-v1\\pilot-v3\\failure.json": "9672c9896978b1e41d6c20cef5471283133760134d38eee24a32ca3b06370e90", |
| "tmp\\select-v3-kompress-readme-comparison-v1\\pilot-v3\\terminal.json": "0f57410a9269b01e8fccc54d71a82855895b2f4c7455b52e24e10511ed478481", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-terminal.json": "5b2c8d958fe383d5c2915af3ab2406ba10fa2f3f64fe9f06f33625eebb6772e3", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\approval.json": "b9b0907856a05b39834f62e6179890a38e6c9b444debb462d7be6cb3cdd049fd", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\result.json": "3f174a23c383438198362551394595845f4434616075568a23361d2e060d63e1", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-v3\\terminal.json": "bac005b33a6e3346e7df4bf3b13917ddb2bd4e25f8a06e47f3819b17a87f2046", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-kompress\\failure.json": "1ecc9a6f8b46f54fa20280da0af3ff189012aab471fe1e1e4150357ff5650117", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\pilot-kompress\\terminal.json": "2f4ee317b100ac811eb8d6863120448fbcc49476fe2b6f167bcf084ffee6ea5d", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\offline-needle-presence.json": "8d84b041ac76796bad8041dfa41ab40d536988c5f160e6b7e86a567c00869ead", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\existing-fix-review.json": "33215e4eda79c1339db819cb9db988802122b6ec33fe5491ea933843a6105153", |
| "tmp\\select-v3-kompress-shared-pilot-v2\\integrity-decision.json": "ae1dce406a3793b664a4ee6842969bab04feeb73821485da992ba95823b8c8ed", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\result.json": "15ea520f4425da6c5c1415adee5bf59d31579f93d1dfbc5c9aaf8a6cc868c7ae", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\launch-terminal.json": "05b7b47c5f18efd81f8b88f169005a1dd1ae2f71cb562495c1e3600104a25920", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\scope.json": "da6a37779086a0885d7b9a76374ceeb56895531ebd9959d1952b42abb0e6da5b", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\approval.json": "0446fd9db773b1bff6445aa9e7221867a7c9be54e718ac5d32f2d45d73db2d0a", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\source.json": "793974d8d451dd782b48f0cae06737e43b013e378624ac4b8a3912256a83c58d", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\measurement-summary.json": "0483578986041660424068a13c2eb747211040b22e085ca583060d6685122e2e", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\result.json": "56bee6763b4b43cfffcd58f02427425ab03d68253d30dd2bba9f65b3d807857f", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\scope.json": "5da79d5c782971553e6988849dc04c19b129b6a24e3004e0adac7f4c467be45e", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\approval.json": "05388625b90b13c5ba4fde4489b017b99b2fb2f90c12b8acca884f682917d159", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\user-approval-receipt.json": "89030324909b1cf34b402a80ff36b53a5d6aa5cfbfd90a5ad758896cfe9c2702", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\launch-terminal.json": "25805c90aae130b34c345badd61dad6e0229a943c70a8a7395bf264c99a594c2", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\live\\v3\\session\\worker-result.json": "2c202c1b5b69239b2d02168723c73f34dc700f9af9e4d3c1cc7e61df518e684a", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\live\\kompress\\session\\worker-result.json": "095e269b3ec6b7cd2dc7d6707844f2d3efd8e6280efddf494aa9dfcdcaa505e2", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\measurement-summary.json": "e7492714499efee67cb8fd427e37a285b91c44d42e09accffaf3f282811ea1d1" |
| }, |
| "publication_hold": true, |
| "prior_failed_shared_pilot": { |
| "status": "INCOMPLETE_TERMINAL_NO_RETRY", |
| "dataset": { |
| "repository": "chopratejas/headroom-datasets", |
| "revision": "2fd099f9f565f1af59e22a6a51671faff6a227f9", |
| "rows": 30, |
| "split": "train", |
| "license": "CC-BY-4.0", |
| "source": "synthetic", |
| "downloaded_bytes": 5950123, |
| "sha256": "ef485e1918b0df66976238e72adb3300196378de244caf9032547a545e6f65f8", |
| "permission": "Explicit USER evaluation-only exception, no training.", |
| "attribution": "Headroom Pilot by chopratejas; https://huggingface.co/datasets/chopratejas/headroom-datasets; CC-BY-4.0 https://creativecommons.org/licenses/by/4.0/" |
| }, |
| "selection": { |
| "dataset": "chopratejas/headroom-datasets", |
| "revision": "2fd099f9f565f1af59e22a6a51671faff6a227f9", |
| "license": "CC-BY-4.0", |
| "permission": "explicit user exception for evaluation only; no training", |
| "selection": "lexicographically smallest id in each of the five categories, frozen before model outcomes", |
| "ids": [ |
| "agentic_loop_001", |
| "code_agent_001", |
| "json_data_001", |
| "logs_001", |
| "rag_search_001" |
| ], |
| "dataset_sha256": "ef485e1918b0df66976238e72adb3300196378de244caf9032547a545e6f65f8", |
| "qualification": false |
| }, |
| "scope_sha256": "0a297cc1c5dbc56167bb7cb3ba31d91034fd9b54c18d4bda77fab194e0077c29", |
| "compression_cells": [ |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "v3", |
| "input_tokens": 41097, |
| "target_tokens": 20548, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "scored_then_failed_before_allocation", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "kompress", |
| "input_tokens": 41097, |
| "target_tokens": 20548, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "v3", |
| "input_tokens": 7141, |
| "target_tokens": 3570, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "kompress", |
| "input_tokens": 7141, |
| "target_tokens": 3570, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "v3", |
| "input_tokens": 11345, |
| "target_tokens": 5672, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "kompress", |
| "input_tokens": 11345, |
| "target_tokens": 5672, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "v3", |
| "input_tokens": 39682, |
| "target_tokens": 19841, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "kompress", |
| "input_tokens": 39682, |
| "target_tokens": 19841, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "v3", |
| "input_tokens": 6435, |
| "target_tokens": 3217, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "kompress", |
| "input_tokens": 6435, |
| "target_tokens": 3217, |
| "output_tokens": null, |
| "valid_budget": null, |
| "status": "unexecuted_after_terminal_error", |
| "score_vector_retained": false, |
| "qualification": false |
| } |
| ], |
| "answer_cells": [ |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "raw", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "v3", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "agentic_loop_001", |
| "category": "agentic_loop", |
| "arm": "kompress", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "raw", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "v3", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "code_agent_001", |
| "category": "code_agent", |
| "arm": "kompress", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "raw", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "v3", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "json_data_001", |
| "category": "json_data", |
| "arm": "kompress", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "raw", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "v3", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "logs_001", |
| "category": "logs", |
| "arm": "kompress", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "raw", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "v3", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| }, |
| { |
| "id": "rag_search_001", |
| "category": "rag_search", |
| "arm": "kompress", |
| "status": "unexecuted_after_terminal_error", |
| "scores": null, |
| "prediction": null, |
| "all_requested_values_correct": null |
| } |
| ], |
| "overall_P_R_F1": null, |
| "scoring_request_reached_allocation": 1, |
| "inferred_completed_V3_windows": 6, |
| "inference_basis": "Frozen packing plan has six windows for case0; persisted traceback reaches subprocess.run after successful score() return. No score tensor/vector output was retained.", |
| "completed_compressed_contexts": 0, |
| "kompress_model_loads": 0, |
| "answerer_requests": 0, |
| "error": { |
| "status": "TERMINAL_NO_RETRY", |
| "error_type": "TypeError", |
| "error": "argument should be a str or an os.PathLike object where __fspath__ returns a str, not 'NoneType'" |
| }, |
| "worker_original_waited": true, |
| "guardian_original_waited": true, |
| "root_cause": "Our audit hook assumed subprocess audit executable was a path. Windows reports None unless executable is supplied explicitly; this is harness failure, not evidence of model quality.", |
| "corrected_proposal": { |
| "path": "tmp\\select-v3-kompress-shared-pilot-v2", |
| "scope_sha256": "835ecd96375ba53d5f88ff8d4d3818815e20d3d2d34942eae3fe288163608594", |
| "status": "AWAITING_FRESH_EXPLICIT_USER_APPROVAL", |
| "change": "One explicit subprocess executable keyword; no data/weights/options/budget changes.", |
| "no_model_self_check": "PASS" |
| }, |
| "training": 0, |
| "downloads": 0 |
| }, |
| "fresh_multi_turn_benchmark": { |
| "status": "FIXED_FIVE_REPEAT_BENCHMARK_COMPLETE_DESCRIPTIVE_ONLY", |
| "user_date": "2026-11-04", |
| "scope_sha256": "5da79d5c782971553e6988849dc04c19b129b6a24e3004e0adac7f4c467be45e", |
| "scope_path": "tmp\\select-v3-multiturn-fixed-repeat-v2", |
| "source_id": "oasst:0a222f36-23d0-405d-8b86-f7d878a98992", |
| "source_sha256": "793974d8d451dd782b48f0cae06737e43b013e378624ac4b8a3912256a83c58d", |
| "input_source_count": 1, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "retention_ratio": 0.5, |
| "conditioning_requests_per_arm": 1, |
| "measured_samples_per_arm": 5, |
| "all_samples_reported": true, |
| "source_input_unchanged": true, |
| "clock": "WARM only: start before fresh source JSON/span grouping, exact o200k count and .25/.5 target derivation; include request serialization/IPC, actual model tokenization/forward, API budget adaptation OR native allocation/recount, native child/relay/guardian wait, delivery and caller output recount. Fixed process/tokenizer/model startup+one case0 warmup disclosed separately, NOT cold cost. No result/source/score cache.", |
| "statistic": "Median of all5 predeclared timed observations per arm; not a population p50/p95.", |
| "samples": { |
| "kompress": [ |
| { |
| "arm": "kompress", |
| "replicate_index": 0, |
| "conditioning": false, |
| "complete_request_seconds": 0.0731253, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240232327700, |
| "end_perf_ns": 21240305453000, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 1, |
| "conditioning": false, |
| "complete_request_seconds": 0.0826347, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240306054800, |
| "end_perf_ns": 21240388689500, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 2, |
| "conditioning": false, |
| "complete_request_seconds": 0.0763315, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240389431700, |
| "end_perf_ns": 21240465763200, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 3, |
| "conditioning": false, |
| "complete_request_seconds": 0.0795998, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240466344800, |
| "end_perf_ns": 21240545944600, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 4, |
| "conditioning": false, |
| "complete_request_seconds": 0.0777508, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240546640900, |
| "end_perf_ns": 21240624391700, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| } |
| ], |
| "v3": [ |
| { |
| "arm": "v3", |
| "replicate_index": 0, |
| "conditioning": false, |
| "complete_request_seconds": 0.073385, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276632357100, |
| "end_perf_ns": 21276705742100, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 1, |
| "conditioning": false, |
| "complete_request_seconds": 0.0682457, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276706401500, |
| "end_perf_ns": 21276774647200, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 2, |
| "conditioning": false, |
| "complete_request_seconds": 0.0665719, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276775212100, |
| "end_perf_ns": 21276841784000, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 3, |
| "conditioning": false, |
| "complete_request_seconds": 0.0628747, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276842343900, |
| "end_perf_ns": 21276905218600, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 4, |
| "conditioning": false, |
| "complete_request_seconds": 0.0658813, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276905819100, |
| "end_perf_ns": 21276971700400, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| } |
| ] |
| }, |
| "timing_slots": [ |
| { |
| "arm": "kompress", |
| "replicate_index": 0, |
| "conditioning": false, |
| "complete_request_seconds": 0.0731253, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240232327700, |
| "end_perf_ns": 21240305453000, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 1, |
| "conditioning": false, |
| "complete_request_seconds": 0.0826347, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240306054800, |
| "end_perf_ns": 21240388689500, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 2, |
| "conditioning": false, |
| "complete_request_seconds": 0.0763315, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240389431700, |
| "end_perf_ns": 21240465763200, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 3, |
| "conditioning": false, |
| "complete_request_seconds": 0.0795998, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240466344800, |
| "end_perf_ns": 21240545944600, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 4, |
| "conditioning": false, |
| "complete_request_seconds": 0.0777508, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21240546640900, |
| "end_perf_ns": 21240624391700, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 0, |
| "conditioning": false, |
| "complete_request_seconds": 0.073385, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276632357100, |
| "end_perf_ns": 21276705742100, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 1, |
| "conditioning": false, |
| "complete_request_seconds": 0.0682457, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276706401500, |
| "end_perf_ns": 21276774647200, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 2, |
| "conditioning": false, |
| "complete_request_seconds": 0.0665719, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276775212100, |
| "end_perf_ns": 21276841784000, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 3, |
| "conditioning": false, |
| "complete_request_seconds": 0.0628747, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276842343900, |
| "end_perf_ns": 21276905218600, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 4, |
| "conditioning": false, |
| "complete_request_seconds": 0.0658813, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276905819100, |
| "end_perf_ns": 21276971700400, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| } |
| ], |
| "conditioning_samples": { |
| "kompress": { |
| "arm": "kompress", |
| "replicate_index": 0, |
| "conditioning": true, |
| "complete_request_seconds": 0.2798597, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 284, |
| "payload_sha256": "e800435aa656b39a9e0a0f99813b9f1958c68a7704a79835967199122f075e17", |
| "clock": { |
| "start_perf_ns": 21239951679500, |
| "end_perf_ns": 21240231539200, |
| "includes_relay_guard_wait": false |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 3, |
| "native_API_calls": 3, |
| "budget_status": "within_target", |
| "native_attempt_tokens": [ |
| 374, |
| 308, |
| 284 |
| ] |
| }, |
| "v3": { |
| "arm": "v3", |
| "replicate_index": 0, |
| "conditioning": true, |
| "complete_request_seconds": 0.3835174, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "output_tokens": 235, |
| "payload_sha256": "4f5b7c564349859e038309697224f682467c2c36b5b2c11430741ea6f769824d", |
| "clock": { |
| "start_perf_ns": 21276248170000, |
| "end_perf_ns": 21276631687400, |
| "includes_relay_guard_wait": true |
| }, |
| "budget_eligible": true, |
| "model_graph_forwards": 1, |
| "maximum_absolute_reference_score_error": 0.0, |
| "scorer_revision": "fd8231d3cdc1efc46f7a6d7d83ca88b336a953591d63ea431407c8b13b990309", |
| "kept_group_ids": [ |
| "831e6a67-fe1d-49d0-af40-ff4a86f6bd7c", |
| "1be03406-5af8-41b8-8c09-fd688bbe942d", |
| "e42fb4d9-a8be-48b6-aa6b-5077c2cedfde" |
| ], |
| "packing_sha256": "f35c67f1ef625c00a2029de58a167a04f6b48b9ffec1bafcd39de75567f23060" |
| } |
| }, |
| "sample_median_seconds": { |
| "kompress": 0.0777508, |
| "v3": 0.0665719 |
| }, |
| "v3_time_change_percent": -14.377858491488183, |
| "matched_budget_eligible": true, |
| "admission_policy": { |
| "gpu_busy_gate_waived": true, |
| "user_instruction": "start now, ignore gate", |
| "original_maximum_gpu_busy_percent": 25, |
| "effective_maximum_gpu_busy_percent": null, |
| "other_resource_and_health_limits_unchanged": true, |
| "record_actual_gpu_load": true, |
| "controlled_idle_benchmark": false |
| }, |
| "profiles": { |
| "kompress": { |
| "setup": { |
| "startup_plus_fixed_warmup_seconds_NOT_cold_request": 39.7928302 |
| }, |
| "model_counts": { |
| "validation": 1, |
| "canary": 1, |
| "compression": 18, |
| "graph_attempted": 20, |
| "graph_completed": 20, |
| "api_calls": 18, |
| "requests": 6, |
| "graph_loads": 1, |
| "parameters": 149999875, |
| "raw_model_loads": 0, |
| "compiler_calls": 0, |
| "BF16_head_projection_observations": 60, |
| "torch_peak_allocated_bytes": 698439680, |
| "torch_peak_reserved_bytes": 748683264 |
| }, |
| "model_lifetime": { |
| "worker_stopped_and_waited": true, |
| "guard_stopped_and_waited": true, |
| "exit_code": 0, |
| "guard_exit_code": 0, |
| "stop_reason": null, |
| "action_completed": true, |
| "action_error_type": null, |
| "intentional_owned_stop_after_action": true, |
| "wall_seconds": 40.53100000000268, |
| "qualification": false, |
| "no_retry": true |
| }, |
| "guard_terminal": { |
| "status": "child_finished", |
| "child_handle_inherited": true, |
| "owner_handle_inherited": true, |
| "stop_target": "original inherited lifetime handle only", |
| "no_retry": true, |
| "worker_pid": 40120, |
| "worker_creation_ticks": 134382999506055568, |
| "guard_pid": 41800, |
| "guard_creation_ticks": 134382999506096688, |
| "child_stopped": true, |
| "child_exit_code": 0 |
| }, |
| "admission": { |
| "index": 0, |
| "start_elapsed_seconds": 0.0, |
| "end_elapsed_seconds": 0.046999999998661224, |
| "query_seconds": 0.046999999998661224, |
| "readiness": { |
| "free_gpu_mib": 12832, |
| "used_gpu_mib": 3146, |
| "gpu_busy_percent": 25, |
| "free_ram_bytes": 8576000000, |
| "free_disk_bytes": 269455609856, |
| "admitted": true, |
| "original_admitted": true, |
| "gpu_busy_gate_waived_by_user": true |
| } |
| } |
| }, |
| "v3": { |
| "setup": { |
| "startup_plus_fixed_warmup_seconds_NOT_cold_request": 35.6139758 |
| }, |
| "model_counts": { |
| "requests": 6, |
| "graph_forwards": 6, |
| "graph_loads": 1, |
| "parameters": 149606401, |
| "raw_model_loads": 0, |
| "compiler_calls": 0, |
| "torch_peak_allocated_bytes": 643732992, |
| "torch_peak_reserved_bytes": 666894336 |
| }, |
| "model_lifetime": { |
| "worker_stopped_and_waited": true, |
| "guard_stopped_and_waited": true, |
| "exit_code": 0, |
| "guard_exit_code": 0, |
| "stop_reason": null, |
| "action_completed": true, |
| "action_error_type": null, |
| "intentional_owned_stop_after_action": true, |
| "wall_seconds": 36.29699999999866, |
| "qualification": false, |
| "no_retry": true |
| }, |
| "guard_terminal": { |
| "status": "child_finished", |
| "child_handle_inherited": true, |
| "owner_handle_inherited": true, |
| "stop_target": "original inherited lifetime handle only", |
| "no_retry": true, |
| "worker_pid": 41140, |
| "worker_creation_ticks": 134382999911848884, |
| "guard_pid": 32980, |
| "guard_creation_ticks": 134382999911896160, |
| "child_stopped": true, |
| "child_exit_code": 0 |
| }, |
| "admission": { |
| "index": 0, |
| "start_elapsed_seconds": 0.0, |
| "end_elapsed_seconds": 0.030999999999039574, |
| "query_seconds": 0.030999999999039574, |
| "readiness": { |
| "free_gpu_mib": 12835, |
| "used_gpu_mib": 3143, |
| "gpu_busy_percent": 38, |
| "free_ram_bytes": 8429604864, |
| "free_disk_bytes": 269455392768, |
| "admitted": true, |
| "original_admitted": false, |
| "gpu_busy_gate_waived_by_user": true |
| } |
| } |
| } |
| }, |
| "model_graph_forwards_total": 26, |
| "raw_model_loads": 0, |
| "compiler_calls": 0, |
| "QA_calls": 0, |
| "model_downloads": 0, |
| "training": 0, |
| "qualification": false, |
| "general_superiority_established": false, |
| "population_quantiles_established": false, |
| "downstream_quality_evaluated": false, |
| "publication_hold": true, |
| "old_historical_row_preserved": true, |
| "Qwen8a_consumed": false, |
| "limitations": [ |
| "Single exposed conversation repeated5 times, not5 independent conversations.", |
| "Serial profile blocks, not interleaved randomized pairs; timing samples are not causal or statistically qualified superiority.", |
| "Only the GPU-busy admission condition was waived by explicit user instruction. Background GPU activity may affect observations; not a controlled-idle comparison.", |
| "The same293-token cap does not imply the same output length or quality: V3 returns235 tokens, Kompress284.", |
| "Process/model startup and fixed conditioning are excluded; setup durations are not cold-request latency.", |
| "No raw dataset text or score vectors are included in this public summary." |
| ], |
| "evidence_sha256": { |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\result.json": "56bee6763b4b43cfffcd58f02427425ab03d68253d30dd2bba9f65b3d807857f", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\scope.json": "5da79d5c782971553e6988849dc04c19b129b6a24e3004e0adac7f4c467be45e", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\approval.json": "05388625b90b13c5ba4fde4489b017b99b2fb2f90c12b8acca884f682917d159", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\user-approval-receipt.json": "89030324909b1cf34b402a80ff36b53a5d6aa5cfbfd90a5ad758896cfe9c2702", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\launch-terminal.json": "25805c90aae130b34c345badd61dad6e0229a943c70a8a7395bf264c99a594c2", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\live\\v3\\session\\worker-result.json": "2c202c1b5b69239b2d02168723c73f34dc700f9af9e4d3c1cc7e61df518e684a", |
| "tmp\\select-v3-multiturn-fixed-repeat-v2\\live\\kompress\\session\\worker-result.json": "095e269b3ec6b7cd2dc7d6707844f2d3efd8e6280efddf494aa9dfcdcaa505e2" |
| } |
| }, |
| "prior_multi_turn_benchmarks": [ |
| { |
| "status": "TERMINAL_PRE_MODEL_RESOURCE_ADMISSION_EXPIRED", |
| "user_date": "2026-11-04", |
| "scope_sha256": "da6a37779086a0885d7b9a76374ceeb56895531ebd9959d1952b42abb0e6da5b", |
| "scope_path": "tmp\\select-v3-multiturn-fixed-repeat-v1", |
| "approval": { |
| "approved": true, |
| "source": "explicit-user-response", |
| "scope_sha256": "da6a37779086a0885d7b9a76374ceeb56895531ebd9959d1952b42abb0e6da5b" |
| }, |
| "user_approval_sha256": "d2359d2bb55e4486f5a463e57db6e0f3c4cb63519ebfe77756d9212c1b4d903c", |
| "source": { |
| "original_source": "E:\\Github\\tokenfold\\tmp\\select-v3-cold-request-dev-v1\\source-inputs\\7.json", |
| "original_case": { |
| "index": 7, |
| "id": "oasst:0a222f36-23d0-405d-8b86-f7d878a98992", |
| "workload": "multi-turn", |
| "ratio": 0.5, |
| "baseline_tokens": 587, |
| "target_tokens": 293, |
| "source_input": "E:\\Github\\tokenfold\\tmp\\select-v3-cold-request-dev-v1\\source-inputs\\7.json", |
| "required_ids": [] |
| }, |
| "selection": "Exact single multi-turn50percent source from requested historical row, repeated5 times; no new easier cases or outcome-dependent replacement.", |
| "new_clock": "Same complete guarded resident request boundary as the latest historical V3 block and retained Kompress block; fresh grouping/count/target through delivery and recount, including applicable relay/guardian wait." |
| }, |
| "source_sha256": "793974d8d451dd782b48f0cae06737e43b013e378624ac4b8a3912256a83c58d", |
| "input_source_count": 1, |
| "input_tokens": 587, |
| "target_tokens": 293, |
| "retention_ratio": 0.5, |
| "conditioning_requests_planned_per_arm": 1, |
| "measured_samples_planned_per_arm": 5, |
| "clock": "WARM only: start before fresh source JSON/span grouping, exact o200k count and .25/.5 target derivation; include request serialization/IPC, actual model tokenization/forward, API budget adaptation OR native allocation/recount, native child/relay/guardian wait, delivery and caller output recount. Fixed process/tokenizer/model startup+one case0 warmup disclosed separately, NOT cold cost. No result/source/score cache.", |
| "statistic": "Median of all5 scheduled measured complete-request observations per arm; primary comparison only when all ten measured cells are valid/budget-met. Do not select minimums or favorable repeats; no adaptive stopping, case replacement, source padding or outcome-based filtering.", |
| "timing_slots": [ |
| { |
| "arm": "kompress", |
| "replicate_index": 0, |
| "status": "not_admitted", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 1, |
| "status": "not_admitted", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 2, |
| "status": "not_admitted", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 3, |
| "status": "not_admitted", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "kompress", |
| "replicate_index": 4, |
| "status": "not_admitted", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 0, |
| "status": "unexecuted_after_first_profile_admission_expired", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 1, |
| "status": "unexecuted_after_first_profile_admission_expired", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 2, |
| "status": "unexecuted_after_first_profile_admission_expired", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 3, |
| "status": "unexecuted_after_first_profile_admission_expired", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| }, |
| { |
| "arm": "v3", |
| "replicate_index": 4, |
| "status": "unexecuted_after_first_profile_admission_expired", |
| "complete_request_seconds": null, |
| "budget_eligible": null, |
| "output_tokens": null |
| } |
| ], |
| "measured_sample_count": 0, |
| "sample_median_seconds": { |
| "kompress": null, |
| "v3": null |
| }, |
| "v3_time_change_percent": null, |
| "model_workers_started": 0, |
| "new_model_forwards": 0, |
| "QA_calls": 0, |
| "model_downloads": 0, |
| "training": 0, |
| "readiness_summary": { |
| "probes": 18, |
| "all_admitted": false, |
| "gpu_busy_limit_percent": 25, |
| "gpu_busy_min_percent": 33, |
| "gpu_busy_max_percent": 38, |
| "maximum_admission_wait_seconds": 180 |
| }, |
| "readiness_probes": [ |
| { |
| "index": 0, |
| "start_elapsed_seconds": 0.0, |
| "end_elapsed_seconds": 0.0459999999984575, |
| "query_seconds": 0.0459999999984575, |
| "readiness": { |
| "free_gpu_mib": 12718, |
| "used_gpu_mib": 3260, |
| "gpu_busy_percent": 36, |
| "free_ram_bytes": 8778215424, |
| "free_disk_bytes": 269457362944, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 1, |
| "start_elapsed_seconds": 10.062000000001717, |
| "end_elapsed_seconds": 10.093000000000757, |
| "query_seconds": 0.030999999999039574, |
| "readiness": { |
| "free_gpu_mib": 12691, |
| "used_gpu_mib": 3287, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 8341643264, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 10, |
| "start_elapsed_seconds": 100.45300000000134, |
| "end_elapsed_seconds": 100.5, |
| "query_seconds": 0.046999999998661224, |
| "readiness": { |
| "free_gpu_mib": 12789, |
| "used_gpu_mib": 3189, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 9655238656, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 11, |
| "start_elapsed_seconds": 110.5, |
| "end_elapsed_seconds": 110.54599999999846, |
| "query_seconds": 0.0459999999984575, |
| "readiness": { |
| "free_gpu_mib": 12789, |
| "used_gpu_mib": 3189, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 9755480064, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 12, |
| "start_elapsed_seconds": 120.54599999999846, |
| "end_elapsed_seconds": 120.57800000000134, |
| "query_seconds": 0.03200000000288128, |
| "readiness": { |
| "free_gpu_mib": 12788, |
| "used_gpu_mib": 3190, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 10181681152, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 13, |
| "start_elapsed_seconds": 130.59300000000076, |
| "end_elapsed_seconds": 130.625, |
| "query_seconds": 0.0319999999992433, |
| "readiness": { |
| "free_gpu_mib": 12734, |
| "used_gpu_mib": 3244, |
| "gpu_busy_percent": 38, |
| "free_ram_bytes": 10179813376, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 14, |
| "start_elapsed_seconds": 140.625, |
| "end_elapsed_seconds": 140.67099999999846, |
| "query_seconds": 0.0459999999984575, |
| "readiness": { |
| "free_gpu_mib": 12721, |
| "used_gpu_mib": 3257, |
| "gpu_busy_percent": 36, |
| "free_ram_bytes": 10120937472, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 15, |
| "start_elapsed_seconds": 150.67099999999846, |
| "end_elapsed_seconds": 150.70300000000134, |
| "query_seconds": 0.03200000000288128, |
| "readiness": { |
| "free_gpu_mib": 12732, |
| "used_gpu_mib": 3246, |
| "gpu_busy_percent": 36, |
| "free_ram_bytes": 10073591808, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 16, |
| "start_elapsed_seconds": 160.70300000000134, |
| "end_elapsed_seconds": 160.75, |
| "query_seconds": 0.046999999998661224, |
| "readiness": { |
| "free_gpu_mib": 12761, |
| "used_gpu_mib": 3217, |
| "gpu_busy_percent": 36, |
| "free_ram_bytes": 10046738432, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 17, |
| "start_elapsed_seconds": 170.75, |
| "end_elapsed_seconds": 170.78099999999904, |
| "query_seconds": 0.030999999999039574, |
| "readiness": { |
| "free_gpu_mib": 12760, |
| "used_gpu_mib": 3218, |
| "gpu_busy_percent": 36, |
| "free_ram_bytes": 10007683072, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 2, |
| "start_elapsed_seconds": 20.093000000000757, |
| "end_elapsed_seconds": 20.139999999999418, |
| "query_seconds": 0.046999999998661224, |
| "readiness": { |
| "free_gpu_mib": 12758, |
| "used_gpu_mib": 3220, |
| "gpu_busy_percent": 35, |
| "free_ram_bytes": 8410136576, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 3, |
| "start_elapsed_seconds": 30.139999999999418, |
| "end_elapsed_seconds": 30.187000000001717, |
| "query_seconds": 0.0470000000022992, |
| "readiness": { |
| "free_gpu_mib": 12738, |
| "used_gpu_mib": 3240, |
| "gpu_busy_percent": 35, |
| "free_ram_bytes": 8980393984, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 4, |
| "start_elapsed_seconds": 40.18700000000172, |
| "end_elapsed_seconds": 40.21800000000076, |
| "query_seconds": 0.030999999999039574, |
| "readiness": { |
| "free_gpu_mib": 12759, |
| "used_gpu_mib": 3219, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 9586401280, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 5, |
| "start_elapsed_seconds": 50.23400000000038, |
| "end_elapsed_seconds": 50.26499999999942, |
| "query_seconds": 0.030999999999039574, |
| "readiness": { |
| "free_gpu_mib": 12754, |
| "used_gpu_mib": 3224, |
| "gpu_busy_percent": 35, |
| "free_ram_bytes": 9567227904, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 6, |
| "start_elapsed_seconds": 60.28099999999904, |
| "end_elapsed_seconds": 60.31200000000172, |
| "query_seconds": 0.031000000002677552, |
| "readiness": { |
| "free_gpu_mib": 12764, |
| "used_gpu_mib": 3214, |
| "gpu_busy_percent": 33, |
| "free_ram_bytes": 9498181632, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 7, |
| "start_elapsed_seconds": 70.31200000000172, |
| "end_elapsed_seconds": 70.35900000000038, |
| "query_seconds": 0.046999999998661224, |
| "readiness": { |
| "free_gpu_mib": 12774, |
| "used_gpu_mib": 3204, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 9478062080, |
| "free_disk_bytes": 269457358848, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 8, |
| "start_elapsed_seconds": 80.35900000000038, |
| "end_elapsed_seconds": 80.40599999999904, |
| "query_seconds": 0.046999999998661224, |
| "readiness": { |
| "free_gpu_mib": 12784, |
| "used_gpu_mib": 3194, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 9646678016, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| }, |
| { |
| "index": 9, |
| "start_elapsed_seconds": 90.40599999999904, |
| "end_elapsed_seconds": 90.43700000000172, |
| "query_seconds": 0.031000000002677552, |
| "readiness": { |
| "free_gpu_mib": 12787, |
| "used_gpu_mib": 3191, |
| "gpu_busy_percent": 34, |
| "free_ram_bytes": 9651068928, |
| "free_disk_bytes": 269457354752, |
| "admitted": false |
| } |
| } |
| ], |
| "conditioning_slots": [ |
| { |
| "arm": "kompress", |
| "status": "unexecuted", |
| "complete_request_seconds": null |
| }, |
| { |
| "arm": "v3", |
| "status": "unexecuted", |
| "complete_request_seconds": null |
| } |
| ], |
| "terminal_result": { |
| "status": "TERMINAL_NO_RETRY", |
| "qualification": false, |
| "superiority_proven": false, |
| "planned_samples_per_arm": 5, |
| "input_sources": 1, |
| "selected_favorable_samples": false, |
| "cold_request_measurements": 0, |
| "consumer_calls": 0, |
| "model_downloads": 0, |
| "Qwen8a_consumed": false, |
| "error_type": "ValueError", |
| "error": "bounded_pre_model_admission_expired", |
| "completed_profiles": [], |
| "recorded_samples": {} |
| }, |
| "controller_lifetime": { |
| "worker_original_waited": true, |
| "guardian_original_waited": true, |
| "worker_exit": 0, |
| "guardian_exit": 0, |
| "stop_reason": null, |
| "action_completed": false, |
| "action_error_type": null, |
| "qualification": false |
| }, |
| "original_historical_row_preserved": true, |
| "no_winner_claim": true, |
| "qualification": false, |
| "superiority_proven": false, |
| "no_retry": true, |
| "publication_hold": true, |
| "Qwen8a_consumed": false, |
| "evidence_sha256": { |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\result.json": "15ea520f4425da6c5c1415adee5bf59d31579f93d1dfbc5c9aaf8a6cc868c7ae", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\launch-terminal.json": "05b7b47c5f18efd81f8b88f169005a1dd1ae2f71cb562495c1e3600104a25920", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\scope.json": "da6a37779086a0885d7b9a76374ceeb56895531ebd9959d1952b42abb0e6da5b", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\approval.json": "0446fd9db773b1bff6445aa9e7221867a7c9be54e718ac5d32f2d45d73db2d0a", |
| "tmp\\select-v3-multiturn-fixed-repeat-v1\\source.json": "793974d8d451dd782b48f0cae06737e43b013e378624ac4b8a3912256a83c58d" |
| } |
| } |
| ] |
| } |
|
|
