194 lines
8.3 KiB
JSON
194 lines
8.3 KiB
JSON
{
|
|
"schema_version": 1,
|
|
"campaign_id": "predator-media-20261021",
|
|
"title": "Images, video and music on a 6 GB GPU",
|
|
"hardware": {
|
|
"gpu": "NVIDIA GeForce GTX 1060 6 GB",
|
|
"ram_gib_approx": 28,
|
|
"os": "Windows 10 Pro 19045",
|
|
"driver": "582.28"
|
|
},
|
|
"runtime": {
|
|
"comfyui": "0.42.1",
|
|
"comfyui_commit": "63be2f9b82ad0dac88a7ceaa28a9cf3154125a18",
|
|
"gguf_node_commit": "6ea2651e7df66d7585f6ffee804b20e92fb38b8a",
|
|
"python": "3.12.10",
|
|
"torch": "2.14.0+cu126",
|
|
"provider": "Local ComfyUI through the Packet Foundry relay",
|
|
"compute": "FP32 unless specific graph explicitly states otherwise; storage quantization is separate.",
|
|
"packet_foundry": {
|
|
"version": "0.1.2",
|
|
"contracts_version": "0.1.2",
|
|
"contracts_wheel_sha256": "732f48490feb87ab45c8cbc93d2c6eb3e0c207cc97fd0c93910a07463db4899b",
|
|
"factory_wheel_sha256": "88efe9fd16c0b238394e686cba8185aba3c1e0bac2ab0c98f424fc9565d9acc5",
|
|
"execution_commit": "Not recorded inthis retainedhardwareproof; currentcheckoutHEAD not used as substitute."
|
|
}
|
|
},
|
|
"coverage": {
|
|
"retained_output_files": 123,
|
|
"accepted_outputs": 122,
|
|
"images": 110,
|
|
"accepted_images": 109,
|
|
"image_complexity_suite": 100,
|
|
"videos": 11,
|
|
"music": 2,
|
|
"contact_sheets": 11,
|
|
"failed_delivery_outputs_with_media": 1,
|
|
"failed_or_cancelled_submitted_jobs_without_media": 7,
|
|
"unique_original_media_sha256": 121,
|
|
"diagnostic_fixture_exclusion": "Two synthetic codec preflight MP4s are fixtures, not model outputs; excluded from gallery and generated-output counts.",
|
|
"duplicate_media_note": "Two pairs of early accepted image files have byte-identical source hashes. Each delivery is retained;123files does not mean123unique contents."
|
|
},
|
|
"methodology": [
|
|
"Same model, template and fixed generation settings; record any intentional prompt change and seed42 then43. A changed prompt forces fresh text encoding; a changed seed forces sampling.",
|
|
"No intentional process restart or model unload between the pair; report what was actually observed.",
|
|
"Do not call a run warm or cold solely from ordering.",
|
|
"Record cached nodes and loading/encoding observations separately from sampler progress.",
|
|
"Any subtraction residual is combined loading, encoding, decoding and other overhead; never exact weight-only loading.",
|
|
"Separate interrupted runs and changed step counts from a matched pair.",
|
|
"Background downloads and prompt/model caching are recorded confounders."
|
|
],
|
|
"summary": {
|
|
"image_complexity": {
|
|
"generated_at": "2026-10-21T01:22:55.914100+00:00",
|
|
"suite_id": "[private identifier]",
|
|
"suite_name": "Predator: fifty progressive image briefs, two consecutive seed variants each",
|
|
"status": "complete_generation_delivery_traces_experiments_gallery",
|
|
"batches": {
|
|
"first": "[private identifier]",
|
|
"second": "[private identifier]"
|
|
},
|
|
"totals": {
|
|
"planned_requests": 100,
|
|
"verified_images": 100,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 100,
|
|
"timed_verified_images": 100,
|
|
"sum_seconds": 6116.5630790000005,
|
|
"median_seconds": 55.9402515,
|
|
"mean_seconds": 61.16563079,
|
|
"p95_seconds": 99.69203990000001,
|
|
"min_seconds": 39.273643,
|
|
"max_seconds": 242.338818
|
|
},
|
|
"briefs": 50,
|
|
"complete_pairs": 50,
|
|
"by_tier": {
|
|
"1": {
|
|
"planned_requests": 20,
|
|
"verified_images": 20,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 20,
|
|
"timed_verified_images": 20,
|
|
"sum_seconds": 1235.124388,
|
|
"median_seconds": 52.97176,
|
|
"mean_seconds": 61.7562194,
|
|
"p95_seconds": 75.45642390000013,
|
|
"min_seconds": 39.273643,
|
|
"max_seconds": 242.338818
|
|
},
|
|
"2": {
|
|
"planned_requests": 20,
|
|
"verified_images": 20,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 20,
|
|
"timed_verified_images": 20,
|
|
"sum_seconds": 1073.9645420000002,
|
|
"median_seconds": 52.956626,
|
|
"mean_seconds": 53.6982271,
|
|
"p95_seconds": 69.7488884,
|
|
"min_seconds": 39.288523,
|
|
"max_seconds": 72.466618
|
|
},
|
|
"3": {
|
|
"planned_requests": 20,
|
|
"verified_images": 20,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 20,
|
|
"timed_verified_images": 20,
|
|
"sum_seconds": 1122.19724,
|
|
"median_seconds": 54.4405185,
|
|
"mean_seconds": 56.109862,
|
|
"p95_seconds": 75.6140475,
|
|
"min_seconds": 39.287177,
|
|
"max_seconds": 75.63889
|
|
},
|
|
"4": {
|
|
"planned_requests": 20,
|
|
"verified_images": 20,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 20,
|
|
"timed_verified_images": 20,
|
|
"sum_seconds": 1231.137535,
|
|
"median_seconds": 60.56643699999999,
|
|
"mean_seconds": 61.55687675,
|
|
"p95_seconds": 81.6720717,
|
|
"min_seconds": 42.349654,
|
|
"max_seconds": 81.696329
|
|
},
|
|
"5": {
|
|
"planned_requests": 20,
|
|
"verified_images": 20,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 20,
|
|
"timed_verified_images": 20,
|
|
"sum_seconds": 1454.139374,
|
|
"median_seconds": 71.1109395,
|
|
"mean_seconds": 72.7069687,
|
|
"p95_seconds": 102.8107459,
|
|
"min_seconds": 45.341639,
|
|
"max_seconds": 102.832423
|
|
}
|
|
},
|
|
"by_arm": {
|
|
"first": {
|
|
"planned_requests": 50,
|
|
"verified_images": 50,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 50,
|
|
"timed_verified_images": 50,
|
|
"sum_seconds": 4058.006349999999,
|
|
"median_seconds": 74.03037850000001,
|
|
"mean_seconds": 81.160127,
|
|
"p95_seconds": 101.4683017,
|
|
"min_seconds": 66.467404,
|
|
"max_seconds": 242.338818
|
|
},
|
|
"second": {
|
|
"planned_requests": 50,
|
|
"verified_images": 50,
|
|
"failed_or_cancelled": 0,
|
|
"trace_verified": 50,
|
|
"timed_verified_images": 50,
|
|
"sum_seconds": 2058.556729000001,
|
|
"median_seconds": 39.397068000000004,
|
|
"mean_seconds": 41.17113458,
|
|
"p95_seconds": 45.39628895,
|
|
"min_seconds": 39.273643,
|
|
"max_seconds": 45.413099
|
|
}
|
|
},
|
|
"warnings": [],
|
|
"limits": [
|
|
"Only succeeded, verified, locally decodable files count as images.",
|
|
"Requested traits and prompt fidelity require visual review; decoding is not a quality score.",
|
|
"Card word_count, prompt and prompt_sha256 fields describe the original brief; the encoder also receives the fixed instruction/source-label envelope.",
|
|
"Matching encoder probe word counts describe the composed text; padded token slots include tokenizer template/padding and may be reused from a cached encoding.",
|
|
"Timings are observed trace duration_seconds values, not a controlled causal comparison.",
|
|
"Prompt detail and constraint count change with tier. Seed42/43 differ across arms.",
|
|
"Cold/warm/cache claims require native runtime and cache evidence; arm labels alone do not establish them."
|
|
]
|
|
},
|
|
"quality": "All100 complexity-image thumbnails showed recognizable scenes; some lettering was malformed. The user described both YuE2 samples as usable music. Video results ranged from nearly static fox scenes and colored noise to recognizable scenes with poor framing and prompt adherence.",
|
|
"failed_jobs_preserved": 7
|
|
},
|
|
"limits": [
|
|
"A valid media container, successful relay, or LangSmith trace is not a quality score.",
|
|
"Model comparisons differ in steps, conditioning, architecture, prompts and cache state. This is practical exploratory evidence, not a controlled model leaderboard.",
|
|
"Sampled resource extrema are not continuous peaks. PhysicalWDDM paging and whole-pipelineGPU residency were not established.",
|
|
"No model weights or private credentials are redistributed. Source hashes identify retained originals; public metadata-stripped copies have separate hashes."
|
|
],
|
|
"archive_run_id": "3cc53336-2555-5e90-99c3-7b94dcb40acc",
|
|
"dataset_sha256": "e67fd94071a63873c156e97e4acd3230b38e4928dfc7c01f01bf79272988c451"
|
|
}
|