Prepare one item1 postfit review with runtime custody and unchanged comparison
This commit is contained in:
1 parent
1c6aeeb566
commit
1f7fccbbfa
28 files changed
+71568
-1
No files matched your search
@@ -3742,6 +3742,26 @@
|
||||
],
|
||||
"review_sha256": "cbd673189c40fe06c1888549ca1c17cfa9d84620c75c42d840e9c614a538c68e",
|
||||
"first_four_cost_gates_PASS": true,
|
||||
"postfit_started": false
|
||||
"postfit_started": false,
|
||||
"publication": "1c6aeeb566026a0df15dfdfacd5765a799420277",
|
||||
"HTTP_readback": "4/4PASS",
|
||||
"publication_receipt": "artifacts/phase3-data-sufficiency-v2-start-publication-2026-11-18/publication.json",
|
||||
"HTTP_receipt": "artifacts/phase3-data-sufficiency-v2-start-publication-2026-11-18/http-readback.json"
|
||||
},
|
||||
"phase3_item1_postfit": {
|
||||
"status": "PREPARED_PENDING_NEW_ASSIST_REVIEW_NOT_LAUNCHED",
|
||||
"root": "/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18",
|
||||
"implementation": "53cca79a47caebaa1ddb8a10e201c1b050dfa258",
|
||||
"contract_sha256": "4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa",
|
||||
"proof_sha256": "0a073cef46a486947b7c0655b722bfda61776ab0d82bf19ac4abd09122f0103d",
|
||||
"binding_sha256": "c196b21571ed901e7a8aa3874b34c53fb8a4ef798e6a5694dd804d153fb391b1",
|
||||
"comparison_sha256": "332f2515911c99aadf2486086ddff82bd195d889faa8391296443e7aeacdfee9",
|
||||
"code_files": 46,
|
||||
"synthetic_checks": 27,
|
||||
"one_review_ruling": "1794950613.195499",
|
||||
"training_untouched": true,
|
||||
"new_predictions": 0,
|
||||
"test_targets_read": false,
|
||||
"unix": 1794951870.1452928
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,390 @@
|
||||
{
|
||||
"parent": "ae089067a17bac613e20441d6c3f8ffa994c06df",
|
||||
"files": [
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-audit-attempt1.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-audit-attempt1.json",
|
||||
"sha256": "f9e8eb33119066545758ede525fc7fb5972e07bc08f615ce1511c01c7248027a",
|
||||
"bytes": 171,
|
||||
"oid": "a65e78289e42d4993d671cdd4f01d4177ccecd01"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-audit.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-audit.json",
|
||||
"sha256": "3159b4b7aa6da1d487015856978eb1cc6e3845ddd71eaa58820d2a914ecea37a",
|
||||
"bytes": 8549,
|
||||
"oid": "d60e0d8df5c3e1fb3d330a54dd3e7749ad142902"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-authorization.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-authorization.json",
|
||||
"sha256": "c0cc2d2f8b12056115c95276c1c7567b89e62ff6f348852d58272fb2877300d7",
|
||||
"bytes": 178,
|
||||
"oid": "9f3abca74965b03cd6f76b2bc0ad1c3ace24ddc6"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-git-freeze.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-git-freeze.json",
|
||||
"sha256": "0bbe877dbafddfc36d0300d3492e7e575c93ee915b6007c8edff26ee9424a6aa",
|
||||
"bytes": 15178,
|
||||
"oid": "c5b0138aec36fb084ce1aed6f146227ed37b966e"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job0.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job0.json",
|
||||
"sha256": "3c8798e5b67e50403c7adfbc468850295579a75e78a00c1a254a62bbfc9bf5f5",
|
||||
"bytes": 474,
|
||||
"oid": "4317381ec631933a8ab2fe6f8198f69ab4dd7b5d"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job1.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job1.json",
|
||||
"sha256": "9f236c24b4e53a1e66ce64aafc2046cbcbb3dc83b6dac670f3450a4ae09537a8",
|
||||
"bytes": 460,
|
||||
"oid": "6e0447f4697171409c099f682348ed83847a9ac3"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job2.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job2.json",
|
||||
"sha256": "4d094e4bb4f6ed217222b32ed5e24ef14703f3bfdd8e61b39e5e0224fca0d2ae",
|
||||
"bytes": 472,
|
||||
"oid": "1ad9cfa739de98ee8fafcb12ef1219c820fc7c5f"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job3.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-snapshots/job3.json",
|
||||
"sha256": "c4a07a7b4e2d1f2d0961947b965f1d9303ac6866cc2dc91bd6fea8a6f225bc43",
|
||||
"bytes": 474,
|
||||
"oid": "d6f63e9a6a743c244480fc64dc24ae381a588447"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/contract.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/contract.json",
|
||||
"sha256": "1bdf2331c5127e03bec160835eda9d04ff333c3f70d5fe12a0629ffa82ec5826",
|
||||
"bytes": 128734,
|
||||
"oid": "6f24d92699b882aba610c2a44b0e412540793cb1"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/first-four-cost-gates.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/first-four-cost-gates.json",
|
||||
"sha256": "25d85cfc64fecb574a6f31f74448e8a9bc44ba70f49e2ec9e86bf09a7d3d6785",
|
||||
"bytes": 5110,
|
||||
"oid": "d486340cc2e0a8ceacb9c356b012750fb693290c"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/implementation-review-authorization.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/implementation-review-authorization.json",
|
||||
"sha256": "bd5ba2c253aa6f8e4a4011c2392d0fb835d2638791ca0983262dd297ddeb4e96",
|
||||
"bytes": 178,
|
||||
"oid": "cedc3fc4ea9e22c40f645e7900a596ab01471984"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/implementation-review-git-freeze.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/implementation-review-git-freeze.json",
|
||||
"sha256": "cd735ddfdf2ec051dea888169d025be408c29002de67619c95a96d393725134a",
|
||||
"bytes": 1870,
|
||||
"oid": "2004c264e78da50f7029c73a706cf83faed4f220"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/dispatch.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/dispatch.json",
|
||||
"sha256": "d94cfb07e4fb94d5dbaf27dd29c43518c55dd15ea045f1bf0b8bc6adc98c1120",
|
||||
"bytes": 365,
|
||||
"oid": "64e6bd6604313789e94d9c2566ed7ee2317e12e1"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/first-epoch-projection.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/first-epoch-projection.json",
|
||||
"sha256": "4f9cc2ad0a3f4bdffa88c4854d037d8fa649cf4750437e67c48cc806e2b140e3",
|
||||
"bytes": 1042,
|
||||
"oid": "1a2a33277dfdfb735f35b4e56bf06229ca7459c9"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/idle-polls.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/idle-polls.json",
|
||||
"sha256": "904dbd01267e441c38d2b69ccc2ff54dc300382fc37f07d448e0c4e6052cf633",
|
||||
"bytes": 275,
|
||||
"oid": "aa5739fd5e553183f86030955620651a21f371ba"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/started.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/started.json",
|
||||
"sha256": "69d3fed8d573f0bea56a2f26a0615deeb1be196bd25514976f6825a25731a6d4",
|
||||
"bytes": 224,
|
||||
"oid": "fbcde1f9fd0db4b66eea1308e51a10d303806777"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/worker/environment.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/worker/environment.json",
|
||||
"sha256": "3795f530c15cf41b0932c287dddb5f34cf8a1e91cd6b03bf2341e445941ede5f",
|
||||
"bytes": 647,
|
||||
"oid": "e8e1c1fe3ce97fd0b414bd4ca7de0f6bb439aadd"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/worker/history/0001.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/worker/history/0001.json",
|
||||
"sha256": "a80145ec2d4c09b328795e5e74667bb1705e01eaf12b451547f43013e8570ca3",
|
||||
"bytes": 567,
|
||||
"oid": "2932abbd0f96a14b689e446cb97f4e034e71e84f"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/worker/standardization.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/0/worker/standardization.json",
|
||||
"sha256": "f5c981bedd8d7774c27f332db1df08c839213df9338b3d00f03d1e775927fce5",
|
||||
"bytes": 7120,
|
||||
"oid": "f51c47acac19da842cc4fa3380b027c0bdb6b579"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/dispatch.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/dispatch.json",
|
||||
"sha256": "79de41f680c869298f3ab0387a3309c6ce68b3b4a621a975bc55fafbe2af17c5",
|
||||
"bytes": 365,
|
||||
"oid": "28891fc9076ca60a71c1906a4bcdae0066fa9af1"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/first-epoch-projection.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/first-epoch-projection.json",
|
||||
"sha256": "db2e56d56c8fd6dc2a57bf64eab9891e73f11fd9d6e634c7d3109169e98090f7",
|
||||
"bytes": 1041,
|
||||
"oid": "fd5e70fcd27214658c797cc1e404b4127cce795a"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/idle-polls.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/idle-polls.json",
|
||||
"sha256": "a63c986f7a015afdc69931dc80a8cd00798020e665fd0bbc6d91520d5af82b9d",
|
||||
"bytes": 276,
|
||||
"oid": "e24bac2dbf9b15d2179dbce82f0b50f2ad0794e7"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/started.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/started.json",
|
||||
"sha256": "63dd0dfbc2485b341b0aa51a740c3f9dda7136e04fcdb1f900b29e2925e656ed",
|
||||
"bytes": 224,
|
||||
"oid": "62eeb9d1cd9210eaca9bd563876f1c28b59117bd"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/worker/environment.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/worker/environment.json",
|
||||
"sha256": "a612f246d03058b3b1599f09b03b56c4c8cdce3e5cca585702d1ccdccba68715",
|
||||
"bytes": 647,
|
||||
"oid": "b375a22f52d9e0966cdc83280103aea35c9a40cf"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/worker/history/0001.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/worker/history/0001.json",
|
||||
"sha256": "6e25e2db8554114984aa7e9a08288423d1886793e1ba793af95e960ecf50df57",
|
||||
"bytes": 574,
|
||||
"oid": "9aa629dcfa47c0a8665fcdd25fcc1a4ef8833621"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/worker/standardization.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/1/worker/standardization.json",
|
||||
"sha256": "c7f8611d911adde3152e4c0b5d2543e9db06796325cb191a0f7efb5147d3d865",
|
||||
"bytes": 11071,
|
||||
"oid": "f6e504862b33ab7dccbbc619dc6381b79d8ea266"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/dispatch.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/dispatch.json",
|
||||
"sha256": "3f1fe4097e19e58192415f9f5178da97e7bece6ebe528661ca903afee8e21f8c",
|
||||
"bytes": 365,
|
||||
"oid": "3e8fbcc2fd7baff51e4f34061693e2d94c4d6472"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/first-epoch-projection.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/first-epoch-projection.json",
|
||||
"sha256": "6259abf484e7ed89dbedf41e74f05d4f632ea2982b70f1c20ef9990fc1943964",
|
||||
"bytes": 1041,
|
||||
"oid": "8f15c373abdd441e183e36456550e0681fbd9a8f"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/idle-polls.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/idle-polls.json",
|
||||
"sha256": "bd2463123f02f97cb9c0b396765a2c4a5d6dee7782bc42ebd0a04b811ffb8619",
|
||||
"bytes": 276,
|
||||
"oid": "bdede40529cbdbff613ce944f35056a54cfdca6f"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/started.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/started.json",
|
||||
"sha256": "a81072edb44f378f40979af6595c5eb1d015125bfffde2d3b3b948baecae2658",
|
||||
"bytes": 224,
|
||||
"oid": "991a25f874d9c160232400be20c27d3b27418378"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/worker/environment.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/worker/environment.json",
|
||||
"sha256": "907398895dc951da0704ba6a524fd889f3747f86f57e9139d0eeaff67e3970cb",
|
||||
"bytes": 647,
|
||||
"oid": "0914331851acb79a642517ae5c949047bcd5dc0c"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/worker/history/0001.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/worker/history/0001.json",
|
||||
"sha256": "5ffceaaef3a0a0f6752d83127f3b6fc38dcdc69d0f4c9fde70ed7f6fb0930372",
|
||||
"bytes": 569,
|
||||
"oid": "dce7baf602d6ded552ac49424bcc85fd21218eea"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/worker/standardization.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/2/worker/standardization.json",
|
||||
"sha256": "6d776bf5e3d23b7995e7f7e84cdc7ec0c7008d7b8fe968400238522c84264c0b",
|
||||
"bytes": 11078,
|
||||
"oid": "79cad53c1a17f650dfff8b4862c8d7e79d7bc973"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/dispatch.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/dispatch.json",
|
||||
"sha256": "5fa6d77879decaf6e09ef0159d223aaf05b1ba3e38c5a2e1772e2a6f919f095b",
|
||||
"bytes": 364,
|
||||
"oid": "2320c173130cba8711a0fec17eca0848e7be29ef"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/first-epoch-projection.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/first-epoch-projection.json",
|
||||
"sha256": "d7934005e8f1b75000d7f3d607ddf51687dde5d631c1e706648a9283471628bd",
|
||||
"bytes": 1042,
|
||||
"oid": "7ecb0cda547f89123babf65ce77807b4983b524e"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/idle-polls.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/idle-polls.json",
|
||||
"sha256": "8db65f854642eaa32f81c21a25b355573fcc3268f630b4d5da6ffb249d5a0598",
|
||||
"bytes": 276,
|
||||
"oid": "b241a551d931a4320d7e8add7bfd41287e81adc8"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/started.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/started.json",
|
||||
"sha256": "2994500838995108860a44e0e532ead4c7bea69c543c5d56e6c5c96ca260052b",
|
||||
"bytes": 224,
|
||||
"oid": "0d57ffc571f3dbeef809d3968a6c7f16253d5d8b"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/worker/environment.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/worker/environment.json",
|
||||
"sha256": "f66cc1908af8881c8cf185a0feb9e29ac642c69e1a10e8eed1f1173ad18303d4",
|
||||
"bytes": 647,
|
||||
"oid": "47d1ee03fd45c979c4c9583694f3f522551874fb"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/worker/history/0001.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/worker/history/0001.json",
|
||||
"sha256": "a5336ac76b249c85362e51f8385d62117a9d4e7f9bc429e5f982a122e0bace4f",
|
||||
"bytes": 567,
|
||||
"oid": "b2ea0e7e0cac0f5e23d3391f41139a342d1b0053"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/worker/standardization.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/jobs/3/worker/standardization.json",
|
||||
"sha256": "cda3e0899c55663a32c3c5956ffe38045dd4b691b8c220262afade99f77d8f63",
|
||||
"bytes": 11078,
|
||||
"oid": "4b3dba07e4b8c4d1cb4754fcad578c4a5582b58c"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/review-acceptance.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/review-acceptance.json",
|
||||
"sha256": "b4bac64d79a5a31764ddff62043f3ed01310376560cfcd9b58c266107bac5690",
|
||||
"bytes": 2068,
|
||||
"oid": "c4b2f346018c33327e6dfcaf48aca443a02748d7"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/review-static-evidence.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/review-static-evidence.json",
|
||||
"sha256": "d46dfa02e161758eab027e7da77eaddde90eebbfe7dcdb062e279cd4b266fd22",
|
||||
"bytes": 214256,
|
||||
"oid": "d7a952bf756de38fd5f93d5b2eb720bfbde1323e"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/review.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/review.json",
|
||||
"sha256": "cbd673189c40fe06c1888549ca1c17cfa9d84620c75c42d840e9c614a538c68e",
|
||||
"bytes": 6490,
|
||||
"oid": "13e285499ee1d57df0619488e650f781a09a8234"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/service-launch-request.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/service-launch-request.json",
|
||||
"sha256": "00f3eb3d423e17630cf05b8b74065a7a4f971ca9216ea86d4acc144a40e62055",
|
||||
"bytes": 1338,
|
||||
"oid": "fcbcee5ed1827bb94cc999fb359b7ee91b3fd706"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/service-launch-result.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/service-launch-result.json",
|
||||
"sha256": "5c0901bdf158018b7b714d191c539122ecceeacc8d9f703d66d89d37fd704cd6",
|
||||
"bytes": 155,
|
||||
"oid": "790d4daabb09673da96b78244f32905a19f7e1f4"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18/started.json",
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/started.json",
|
||||
"sha256": "74fb3adffd76b57fff3b55aff6774cd4a617ff5a4ed0ae4e5aacb9d5b117c78f",
|
||||
"bytes": 321,
|
||||
"oid": "e5ee455df7624ec62e7c522d764a323a32e90a05"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/artifacts/meltassist-phase3-item1-v2-review-2026-11-18/review.json",
|
||||
"path": "artifacts/meltassist-phase3-item1-v2-review-2026-11-18/review.json",
|
||||
"sha256": "cbd673189c40fe06c1888549ca1c17cfa9d84620c75c42d840e9c614a538c68e",
|
||||
"bytes": 6490,
|
||||
"oid": "13e285499ee1d57df0619488e650f781a09a8234"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/artifacts/meltassist-phase3-item1-v2-review-2026-11-18/static-evidence.json",
|
||||
"path": "artifacts/meltassist-phase3-item1-v2-review-2026-11-18/static-evidence.json",
|
||||
"sha256": "d46dfa02e161758eab027e7da77eaddde90eebbfe7dcdb062e279cd4b266fd22",
|
||||
"bytes": 214256,
|
||||
"oid": "d7a952bf756de38fd5f93d5b2eb720bfbde1323e"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/artifacts/phase2-current-handoff-2026-11-17.json",
|
||||
"path": "artifacts/phase2-current-handoff-2026-11-17.json",
|
||||
"sha256": "4eb00b915ccf88c658320e4ca3d9da0bf9398a990d537b53111665409a37c347",
|
||||
"bytes": 134856,
|
||||
"oid": "e7b86b487f0f45507b186beb9a7abe0e7b0c9948"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/artifacts/phase3-data-sufficiency-v2-entry-publication-2026-11-18/file-map.json",
|
||||
"path": "artifacts/phase3-data-sufficiency-v2-entry-publication-2026-11-18/file-map.json",
|
||||
"sha256": "0752fda96f4162242bc643787f418e0f227c186ce302f175770d554c8d507fe6",
|
||||
"bytes": 13952,
|
||||
"oid": "c2e182e6900f0c4895c87646b488d8bd8cf16e81"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/artifacts/phase3-data-sufficiency-v2-entry-publication-2026-11-18/http-readback.json",
|
||||
"path": "artifacts/phase3-data-sufficiency-v2-entry-publication-2026-11-18/http-readback.json",
|
||||
"sha256": "447668d98e86587ee6e5f32308700a3d0c4039904b3d15f79ad221eb0d2e4099",
|
||||
"bytes": 1445,
|
||||
"oid": "4cae2c42d59d5caf2b12121e3722e433415c5717"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/artifacts/phase3-data-sufficiency-v2-entry-publication-2026-11-18/publication.json",
|
||||
"path": "artifacts/phase3-data-sufficiency-v2-entry-publication-2026-11-18/publication.json",
|
||||
"sha256": "7156f6f94fa893c96b2aaddb3e1e4159053ffe1b6c68e4999c8defd24415df32",
|
||||
"bytes": 378,
|
||||
"oid": "b695c36e9bc2441e0f2570bdd09ab693106548d9"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/docs/collaboration-stage-review-policy.md",
|
||||
"path": "docs/collaboration-stage-review-policy.md",
|
||||
"sha256": "f4157869e96831233125ed68c6a657e62de29859c9c70dc7aba2620492a2f59b",
|
||||
"bytes": 96756,
|
||||
"oid": "d2a8405b1cd85d7cf204342644e98aa03327c64f"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/docs/phase2-live-handoff-2026-11-17.md",
|
||||
"path": "docs/phase2-live-handoff-2026-11-17.md",
|
||||
"sha256": "228cb4d7f1b9840966e37b499cc59506f18080300a35c8b698a2a937bf77964e",
|
||||
"bytes": 62575,
|
||||
"oid": "2afc7e68301a198146d0f54cb0e18dc62877bad1"
|
||||
},
|
||||
{
|
||||
"source": "/home/frank14f/optane/MeltMind/docs/phase3-data-sufficiency-v2-actual-start-2026-11-18.md",
|
||||
"path": "docs/phase3-data-sufficiency-v2-actual-start-2026-11-18.md",
|
||||
"sha256": "c4b8994105b2e63631b4c734f13eb51eab28b603678e09ddfa65e2b9fd66084a",
|
||||
"bytes": 2516,
|
||||
"oid": "05ed8cdab90ef777901ac69e58e8857d71d452d4"
|
||||
}
|
||||
]
|
||||
}
|
||||
@@ -0,0 +1,39 @@
|
||||
{
|
||||
"commit": "1c6aeeb566026a0df15dfdfacd5765a799420277",
|
||||
"status": "PASS",
|
||||
"checked_files": [
|
||||
{
|
||||
"path": "docs/phase3-data-sufficiency-v2-actual-start-2026-11-18.md",
|
||||
"http_status": 200,
|
||||
"sha256": "c4b8994105b2e63631b4c734f13eb51eab28b603678e09ddfa65e2b9fd66084a",
|
||||
"expected": "c4b8994105b2e63631b4c734f13eb51eab28b603678e09ddfa65e2b9fd66084a",
|
||||
"matched": true,
|
||||
"bytes": 2516
|
||||
},
|
||||
{
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/actual-start-audit.json",
|
||||
"http_status": 200,
|
||||
"sha256": "3159b4b7aa6da1d487015856978eb1cc6e3845ddd71eaa58820d2a914ecea37a",
|
||||
"expected": "3159b4b7aa6da1d487015856978eb1cc6e3845ddd71eaa58820d2a914ecea37a",
|
||||
"matched": true,
|
||||
"bytes": 8549
|
||||
},
|
||||
{
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/review.json",
|
||||
"http_status": 200,
|
||||
"sha256": "cbd673189c40fe06c1888549ca1c17cfa9d84620c75c42d840e9c614a538c68e",
|
||||
"expected": "cbd673189c40fe06c1888549ca1c17cfa9d84620c75c42d840e9c614a538c68e",
|
||||
"matched": true,
|
||||
"bytes": 6490
|
||||
},
|
||||
{
|
||||
"path": "home-runs/phase3-data-sufficiency-v2-2026-11-18/first-four-cost-gates.json",
|
||||
"http_status": 200,
|
||||
"sha256": "25d85cfc64fecb574a6f31f74448e8a9bc44ba70f49e2ec9e86bf09a7d3d6785",
|
||||
"expected": "25d85cfc64fecb574a6f31f74448e8a9bc44ba70f49e2ec9e86bf09a7d3d6785",
|
||||
"matched": true,
|
||||
"bytes": 5110
|
||||
}
|
||||
],
|
||||
"unix": 1794949459.716047
|
||||
}
|
||||
@@ -0,0 +1,11 @@
|
||||
{
|
||||
"status": "REMOTE_REF_VERIFIED",
|
||||
"commit": "1c6aeeb566026a0df15dfdfacd5765a799420277",
|
||||
"parent": "ae089067a17bac613e20441d6c3f8ffa994c06df",
|
||||
"branch": "codex/phase2-small-evidence-20261117",
|
||||
"files": 55,
|
||||
"bytes": 972536,
|
||||
"writer_head_unchanged": "95ef2592b91f437f5f3274be8599dc9e5139dba9",
|
||||
"wall_seconds": 10.06151270866394,
|
||||
"raw_bulk_not_included": true
|
||||
}
|
||||
@@ -544,3 +544,10 @@ Claude1794948643将e24共同轨迹改为U/2U独立fresh完整日程,8710最终
|
||||
|
||||
Sole service meltmind-phase3-data-sufficiency-v2-20261118/coordinator993672;真实worker993725/993719/993709/993724,GPU0–3,id4/5待对应GPU释放。根/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18;源73f94bbc/contract1bdf2331不变,实际新PASScbd67318已d7e33593固定并validate。首四完整epoch成本门全部PASS,job3投影105177.934464/120000GPU-s(含pending和7200postfit);id4/5首门未来才执行。无重复启动、无S2重启、旧v1未运行;六fit末权重冻结后停待独立checkpoint-bound postfit审阅。详见docs/phase3-data-sufficiency-v2-actual-start-2026-11-18.md与actual-start-audit.json/first-four-cost-gates.json;以真实最新回执为准,不因idle重启。guard旧s2_started命名按Assist解释为陈旧字段,不能当S2新启动。
|
||||
|
||||
|
||||
## item1 postfit一次审阅包已固定,未启动(2026-11-18)
|
||||
|
||||
Claude1794950613.195499明确一次新代码/比较合同审阅;本条替代旧“实际权重齐后再审”的两阶段措辞。唯一准备根/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18,implementation53cca79a47caebaa1ddb8a10e201c1b050dfa258/contract4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa/proof0a073cef46a486947b7c0655b722bfda61776ab0d82bf19ac4abd09122f0103d;bindingc196b21571ed901e7a8aa3874b34c53fb8a4ef798e6a5694dd804d153fb391b1。46code pins/27CPU合成PASS,comparison332f2515字节保持。等待新的Assist PASS,不复用训练PASS;尚未启动postfit/waiter、无预测/评分/目标数组读取。
|
||||
|
||||
运行时先六训练PASS+models/terminal固定+进程退出,实际checkpoint与normalization核验/固定,再7端点及3clock共4582条预测清单冻结,再核同一PASS和两冻结清单后单次读552targets(test121)。S1原121test预测/指标按hash复用;6新训练endpoint不伪造提前hash。postfit7200GPU-s/3600CPU-s在item120000内,原训练旧账一次;不新增fit/retry/后续阶段。原993672训练及73f94bbc/1bdf2331保持。详见docs/phase3-item1-postfit-2026-11-18.md和root/preparation-audit.json,历史启动或下载话题不得重开。
|
||||
|
||||
@@ -1,3 +1,11 @@
|
||||
## item1 postfit一次审阅包已固定,未启动(2026-11-18)
|
||||
|
||||
Claude1794950613.195499明确一次新代码/比较合同审阅;本条替代旧“实际权重齐后再审”的两阶段措辞。唯一准备根/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18,implementation53cca79a47caebaa1ddb8a10e201c1b050dfa258/contract4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa/proof0a073cef46a486947b7c0655b722bfda61776ab0d82bf19ac4abd09122f0103d;bindingc196b21571ed901e7a8aa3874b34c53fb8a4ef798e6a5694dd804d153fb391b1。46code pins/27CPU合成PASS,comparison332f2515字节保持。等待新的Assist PASS,不复用训练PASS;尚未启动postfit/waiter、无预测/评分/目标数组读取。
|
||||
|
||||
运行时先六训练PASS+models/terminal固定+进程退出,实际checkpoint与normalization核验/固定,再7端点及3clock共4582条预测清单冻结,再核同一PASS和两冻结清单后单次读552targets(test121)。S1原121test预测/指标按hash复用;6新训练endpoint不伪造提前hash。postfit7200GPU-s/3600CPU-s在item120000内,原训练旧账一次;不新增fit/retry/后续阶段。原993672训练及73f94bbc/1bdf2331保持。详见docs/phase3-item1-postfit-2026-11-18.md和root/preparation-audit.json,历史启动或下载话题不得重开。
|
||||
|
||||
最新实际启动交付:1c6aeeb566026a0df15dfdfacd5765a799420277已远端ref及4HTTP/SHA核验;含原PASScbd67318、actual-start及首四成本门。仅此一个item1 v2队列993672继续,其他旧启动快照均不能重开。
|
||||
|
||||
## item1 v2唯一实际启动,首四门PASS(2026-11-18)
|
||||
|
||||
Sole service meltmind-phase3-data-sufficiency-v2-20261118/coordinator993672;真实worker993725/993719/993709/993724,GPU0–3,id4/5待对应GPU释放。根/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18;源73f94bbc/contract1bdf2331不变,实际新PASScbd67318已d7e33593固定并validate。首四完整epoch成本门全部PASS,job3投影105177.934464/120000GPU-s(含pending和7200postfit);id4/5首门未来才执行。无重复启动、无S2重启、旧v1未运行;六fit末权重冻结后停待独立checkpoint-bound postfit审阅。详见docs/phase3-data-sufficiency-v2-actual-start-2026-11-18.md与actual-start-audit.json/first-four-cost-gates.json;以真实最新回执为准,不因idle重启。guard旧s2_started命名按Assist解释为陈旧字段,不能当S2新启动。
|
||||
|
||||
@@ -0,0 +1,31 @@
|
||||
# Phase3 item1: one-review postfit binding
|
||||
|
||||
Claude1794950613.195499 supersedes the earlier two-review wording: MeltAssist reviews this code and the unchanged comparison contract once, while training continues. PASS binds the implementation, contract, implementation-proof, comparison and complete source map. Future checkpoint hashes are not invented and the training PASS cannot substitute. This preparation launches no predictor, scorer or waiter and reads no target arrays.
|
||||
|
||||
Entry: `tools/phase3_item1_postfit.py /home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18`, using the original reproduction venv and project working directory. The root's expected-review-binding.json supplies exact identities after preparation. Missing review or incomplete training refuses before a started marker. There is one attempt, no retry or automatic next phase.
|
||||
|
||||
## Runtime custody
|
||||
|
||||
All six original v2 fits must be PASS with zero exit codes, unique job/model identities, frozen models and terminal receipts, no global stop, and no queue/guard/worker processes. Their guard hashes, worker-result hashes, final checkpoint/normalization hashes, roles and steps are reconciled. Original training GPU costs are summed once and must match the settled ledger. The actual six checkpoints and the historical S1 endpoint are then adopted and frozen. Tensor metadata checks enforce fold0, seed0, pool, complete-schedule endpoint, steps and original training contract before prediction.
|
||||
|
||||
Seven endpoint jobs run serially on original GPU0 after three idle checks. Each uses only its own learning set plus the common111 validation and121 test geometries, original full-graph FP32 model and FP64 clock decoding. No AMP/TF32, target branch, normalization fit, clipping or model fit is introduced. The historical S1 arm reuses all121 original test prediction hashes and stored metrics; only its learning/validation predictions are new. Three original learning-only clock fields are also emitted.
|
||||
|
||||
Expected manifest:3265 model entries (3144 new,121 reused) and1317 clock entries =4582 total, covering552 unique parts. Complete arm/pool/domain membership, component identity, node counts, file hashes and exact S1 provenance are checked. The prediction manifest and reviewed code are frozen before any target access. The scorer then revalidates the same exact PASS, actual model freeze and prediction freeze; no second review stage is required.
|
||||
|
||||
Only then is each cached target array read once:320 learning,111 validation,121 test. Target-access records precede each read; they are application evidence, not syscall-level tracing. Existing S1 test metrics are reused unchanged. All new metrics retain the original fixed PPS query, multiplicity once, minimum raw-ID tie convention, query argmax last-location/Rg and original tau pairing. Full-node predictions and negative full/query counts are retained without clipping. There is no fresh VTU/label decoding.
|
||||
|
||||
## Comparisons and interpretation
|
||||
|
||||
Comparison contract is byte-identical to original SHA332f2515911c99aadf2486086ddff82bd195d889faa8391296443e7aeacdfee9. Nested learning pools have34/65/132 components. Slopes use their actual log2 ratios. Common validation/test report s_U,s_2U,g50,g100,s_2U−s_U, epoch slopes, clock slopes and endpoint-minus-clock contrasts. All contrasts share2000 evaluation-component bootstrap draws, seed2026111408. Each component requires every constituent metric defined; missing support and excluded parts/components remain explicit. Own-learning summaries do not subtract metrics across different training domains.
|
||||
|
||||
Field classification uses s_2U and g100: strict CI low>0.001 means benefit, an entire CI within[−0.001,+0.001] means local equivalence; wide/crossing intervals or incomplete support remain unidentified. The four complete-support combinations retain data benefit, update-and-schedule benefit, both, or local plateau. The same observed fold is descriptive, not independent generalization confirmation. Component bootstrap does not measure training seed or pool uncertainty; update effects include complete LR schedule changes. No3U/extrapolation, validation selection, topology closure or next-stage inference is added.
|
||||
|
||||
## Resource limits
|
||||
|
||||
The original item cap is120000 GPU-seconds; training occupies at most112800 and postfit at most7200. Historical S1 GPU cost is not charged again. Each of seven serial endpoint jobs is limited to900 charged GPU-seconds including60 setup and idle polls, and300 CPU-seconds (240 worker plus60 guard reserve). Setup/inter-job wall time is also included conservatively. No automatic retry or alternate GPU selection follows failure. Each enforces15GiB allocated/reserved/device VRAM,16GiB RSS and20GiB output; failed-worker peak records remain checked.
|
||||
|
||||
Postfit CPU cap3600 includes180 setup reserve, coordinator CPU and reaped child CPU,900 scoring reserve and120 terminal reserve. Remaining-job reserves are checked before dispatch. Scoring has900 CPU-seconds/1200 wall-seconds/16GiB RSS. Original training costs are kept separate and counted once; final costs include terminal-freeze work. Resource failure preserves receipts and stops.
|
||||
|
||||
## Acceptance and scope
|
||||
|
||||
Initial18 tests passed; retained attempts2/3 have27 PASS each (attempt3 includes the runtime original-training source-pin check), adding six isolated actual Git/file review-validator checks and three custody/terminal tests. Checks cover strict/inclusive CI boundaries, missing whole components, common bootstrap cancellation, actual count denominator, role rejection before I/O, label-free prediction structure, checkpoint selection, pending training, budget reservations, retained failed VRAM peaks, exact review/source mutations, and rejection before target access. These are CPU synthetic checks, not real prediction/timing validation. Exact resource feasibility remains enforced at runtime. Live training source pins are checked read-only; no training source is modified.
|
||||
@@ -0,0 +1,9 @@
|
||||
{
|
||||
"status": "PASS",
|
||||
"tests": 27,
|
||||
"elapsed_seconds": 1.317,
|
||||
"log_sha256": "6c452434cfe3d4fbde4ce6a9278b6c49c111c603dd2deabc5a24da33d380269a",
|
||||
"test_source_sha256": "7b2187d54a611924e66bc64cfcc7027f421c05a08eb7744889de64a8feba05ba",
|
||||
"scope": "CPU synthetic and isolated Git/file validator; no real targets, prediction, fits or CUDA workload",
|
||||
"original18_pass_preserved": true
|
||||
}
|
||||
@@ -0,0 +1,306 @@
|
||||
{
|
||||
"spec_sha256": "e24fff7c5cb91dff7acf1532f502daec60e4b7fec21b2d4d4dcde1e793be7fb4",
|
||||
"fold": 0,
|
||||
"seed": 0,
|
||||
"pools": [
|
||||
25,
|
||||
50,
|
||||
100
|
||||
],
|
||||
"common_validation": [
|
||||
"00000111_ce24e0cbec8e45c89735d148_step_003",
|
||||
"00000115_ce24e0cbec8e45c89735d148_step_007",
|
||||
"00000116_ce24e0cbec8e45c89735d148_step_008",
|
||||
"00000117_ce24e0cbec8e45c89735d148_step_009",
|
||||
"00000119_ce24e0cbec8e45c89735d148_step_011",
|
||||
"00000120_ce24e0cbec8e45c89735d148_step_012",
|
||||
"00000131_945244874ed84e599987daa9_step_000",
|
||||
"00000132_438507a38e404bfdbd51a16b_step_000",
|
||||
"00000133_08b739e0511f4305b161cd05_step_000",
|
||||
"00000190_185ecbc783f34408aef52119_step_000",
|
||||
"00000196_c9b71ecdd0654716891b0db8_step_000",
|
||||
"00000206_27417003e762409185cb2b69_step_000",
|
||||
"00000238_4829bc0645d6414d9b5cec31_step_001",
|
||||
"00000241_4829bc0645d6414d9b5cec31_step_004",
|
||||
"00000243_4829bc0645d6414d9b5cec31_step_006",
|
||||
"00000248_e87554b09cdc48c3add39db9_step_000",
|
||||
"00000257_845fb06d7f2f4802a806a600_step_002",
|
||||
"00000258_845fb06d7f2f4802a806a600_step_003",
|
||||
"00000259_845fb06d7f2f4802a806a600_step_004",
|
||||
"00000261_845fb06d7f2f4802a806a600_step_006",
|
||||
"00000262_845fb06d7f2f4802a806a600_step_007",
|
||||
"00000264_845fb06d7f2f4802a806a600_step_009",
|
||||
"00000265_845fb06d7f2f4802a806a600_step_010",
|
||||
"00000266_845fb06d7f2f4802a806a600_step_011",
|
||||
"00000267_845fb06d7f2f4802a806a600_step_012",
|
||||
"00000269_845fb06d7f2f4802a806a600_step_014",
|
||||
"00000391_fd490b564a44413884fab719_step_000",
|
||||
"00000405_f3759bbb4682423caf207e8e_step_002",
|
||||
"00000406_f3759bbb4682423caf207e8e_step_003",
|
||||
"00000444_4ed4c78d6d754aac90876fc2_step_003",
|
||||
"00000445_4ed4c78d6d754aac90876fc2_step_004",
|
||||
"00000501_5499914343bd4f549b38046d_step_002",
|
||||
"00000502_5499914343bd4f549b38046d_step_003",
|
||||
"00000574_5a4cef68211d4706b1ec8586_step_003",
|
||||
"00000578_5a4cef68211d4706b1ec8586_step_007",
|
||||
"00000579_5a4cef68211d4706b1ec8586_step_008",
|
||||
"00000580_5a4cef68211d4706b1ec8586_step_009",
|
||||
"00000582_5a4cef68211d4706b1ec8586_step_011",
|
||||
"00000583_5a4cef68211d4706b1ec8586_step_012",
|
||||
"00000597_7fe894812c4244e18241aa08_step_003",
|
||||
"00000601_7fe894812c4244e18241aa08_step_007",
|
||||
"00000602_7fe894812c4244e18241aa08_step_008",
|
||||
"00000603_7fe894812c4244e18241aa08_step_009",
|
||||
"00000605_7fe894812c4244e18241aa08_step_011",
|
||||
"00000606_7fe894812c4244e18241aa08_step_012",
|
||||
"00000654_d87d11d9c3ca4ee5a88797ae_step_002",
|
||||
"00000758_8ab50540e9804365a8717ec0_step_005",
|
||||
"00000759_8ab50540e9804365a8717ec0_step_006",
|
||||
"00000760_8ab50540e9804365a8717ec0_step_007",
|
||||
"00000763_8ab50540e9804365a8717ec0_step_010",
|
||||
"00000816_36e6258346004cabb4c091b2_step_004",
|
||||
"00000817_36e6258346004cabb4c091b2_step_005",
|
||||
"00000892_be26dc83f78947738b2bfabd_step_001",
|
||||
"00000894_be26dc83f78947738b2bfabd_step_003",
|
||||
"00000897_be26dc83f78947738b2bfabd_step_006",
|
||||
"00000919_cdcebf32644443b9812097ba_step_003",
|
||||
"00000923_cdcebf32644443b9812097ba_step_007",
|
||||
"00000927_cdcebf32644443b9812097ba_step_011",
|
||||
"00000928_cdcebf32644443b9812097ba_step_012",
|
||||
"00000955_a741c7966d2c40a8a511263e_step_001",
|
||||
"00000984_0dd1386597104265ab67e6cf_step_007",
|
||||
"00000986_0dd1386597104265ab67e6cf_step_009",
|
||||
"00000987_0dd1386597104265ab67e6cf_step_010",
|
||||
"00000988_0dd1386597104265ab67e6cf_step_011",
|
||||
"00000996_6d71695ba3fd475c927a00a7_step_000",
|
||||
"00001019_ca4cf9693d6d4a98b515ba99_step_000",
|
||||
"00001020_ca4cf9693d6d4a98b515ba99_step_001",
|
||||
"00001021_ca4cf9693d6d4a98b515ba99_step_002",
|
||||
"00001026_0f133d8e7fd448708a554e28_step_000",
|
||||
"00001027_0f133d8e7fd448708a554e28_step_001",
|
||||
"00001040_74c42207fa764831b03f7658_step_000",
|
||||
"00001052_2fd6155fa34f418083d0bd0f_step_001",
|
||||
"00001053_2fd6155fa34f418083d0bd0f_step_002",
|
||||
"00001097_256ee7ec3a4b4972a8b87c8d_step_001",
|
||||
"00001099_256ee7ec3a4b4972a8b87c8d_step_003",
|
||||
"00001102_256ee7ec3a4b4972a8b87c8d_step_006",
|
||||
"00001107_78566795c3144719804f8d34_step_000",
|
||||
"00001167_107f61d0402f4c9a9d9d1342_step_000",
|
||||
"00001168_107f61d0402f4c9a9d9d1342_step_001",
|
||||
"00001217_e517c61f1c49433bb81cf980_step_000",
|
||||
"00001238_96503b6a26c64074ad598927_step_000",
|
||||
"00001239_96503b6a26c64074ad598927_step_001",
|
||||
"00001240_96503b6a26c64074ad598927_step_002",
|
||||
"00001241_96503b6a26c64074ad598927_step_003",
|
||||
"00001242_96503b6a26c64074ad598927_step_004",
|
||||
"00001243_96503b6a26c64074ad598927_step_005",
|
||||
"00001244_96503b6a26c64074ad598927_step_006",
|
||||
"00001246_96503b6a26c64074ad598927_step_008",
|
||||
"00001247_96503b6a26c64074ad598927_step_009",
|
||||
"00001249_96503b6a26c64074ad598927_step_011",
|
||||
"00001252_96503b6a26c64074ad598927_step_014",
|
||||
"00001253_96503b6a26c64074ad598927_step_015",
|
||||
"00001314_a9ed5e56a6a8414f91b39bdb_step_000",
|
||||
"00001333_beab79941e884455a62af40c_step_001",
|
||||
"00001334_beab79941e884455a62af40c_step_002",
|
||||
"00001335_beab79941e884455a62af40c_step_003",
|
||||
"00001336_beab79941e884455a62af40c_step_004",
|
||||
"00001338_beab79941e884455a62af40c_step_006",
|
||||
"00001339_beab79941e884455a62af40c_step_007",
|
||||
"00001340_beab79941e884455a62af40c_step_008",
|
||||
"00001352_4e1d56bc2cef49ea84626b11_step_000",
|
||||
"00001353_4e1d56bc2cef49ea84626b11_step_001",
|
||||
"00001354_4e1d56bc2cef49ea84626b11_step_002",
|
||||
"00001379_5b1d6e2599ea47d498a04256_step_000",
|
||||
"00001387_60cf3ad636dd42e0bcec6946_step_000",
|
||||
"00001511_5cbe58077e4b4885a5c213b3_step_004",
|
||||
"00001512_5cbe58077e4b4885a5c213b3_step_005",
|
||||
"00001525_e0d35a03b46745259cf8878f_step_000",
|
||||
"00001532_9f0a4a641e8f4cc588eb4315_step_000",
|
||||
"00001552_b0a9c4d836294846891615fe_step_000",
|
||||
"00001553_adfa1c33ae4d4d0aa441ddc3_step_000"
|
||||
],
|
||||
"common_test": [
|
||||
"00000006_d4fe04f0f5f84b52bd4f10e4_step_001",
|
||||
"00000008_9b3d6a97e8de4aa193b81000_step_000",
|
||||
"00000009_9b3d6a97e8de4aa193b81000_step_001",
|
||||
"00000011_e909f412cda24521865fac0f_step_000",
|
||||
"00000102_5ed74bccca6f4e89829bcb5e_step_002",
|
||||
"00000183_6c4172e8a17140f89d3b1c0a_step_000",
|
||||
"00000220_5324c79b172f41e8a49afc6f_step_007",
|
||||
"00000221_5324c79b172f41e8a49afc6f_step_008",
|
||||
"00000249_07a2da801a60492fb1e5eeb3_step_000",
|
||||
"00000408_36bcb4030e36417b8dc74b87_step_000",
|
||||
"00000410_30afd12eede44fa8a1def049_step_000",
|
||||
"00000411_305e45c9479f43979b0d7b53_step_000",
|
||||
"00000413_94f421b1b33d47c592c9fe1d_step_000",
|
||||
"00000421_bd0972033f464f8f961a150a_step_002",
|
||||
"00000508_d1c0a44331fd41faa91d4c2e_step_000",
|
||||
"00000509_d1c0a44331fd41faa91d4c2e_step_001",
|
||||
"00000510_d1c0a44331fd41faa91d4c2e_step_002",
|
||||
"00000522_a56dd4152b7e4ccf91212d7b_step_000",
|
||||
"00000527_db75f3a38d9746a1bd66168d_step_000",
|
||||
"00000542_4b1756f6e72b45c29a8e4eb4_step_000",
|
||||
"00000549_41a24c0b74a34217832f5af0_step_000",
|
||||
"00000551_41a24c0b74a34217832f5af0_step_002",
|
||||
"00000562_1724088724bf4e53a36e363e_step_000",
|
||||
"00000566_2f2b015ef4904c1caef56154_step_000",
|
||||
"00000637_5ae45a997c76423aba2fcc11_step_002",
|
||||
"00000638_5ae45a997c76423aba2fcc11_step_003",
|
||||
"00000643_bb55eaa4e9e24c2ea9d18ee0_step_000",
|
||||
"00000659_6fae5591b3034f949b7e8eae_step_000",
|
||||
"00000695_3f46763af0744ab2a90e6b48_step_000",
|
||||
"00000746_6cb3ddbb49974d7db0e97135_step_001",
|
||||
"00000796_dec80701adc84239b485449d_step_002",
|
||||
"00000806_05cadeb104be4c8aa75ccc71_step_000",
|
||||
"00000807_05cadeb104be4c8aa75ccc71_step_001",
|
||||
"00000808_05cadeb104be4c8aa75ccc71_step_002",
|
||||
"00000809_05cadeb104be4c8aa75ccc71_step_003",
|
||||
"00000840_2c1efcc32d094e9fb9e1a88e_step_000",
|
||||
"00000842_2c1efcc32d094e9fb9e1a88e_step_002",
|
||||
"00000843_2c1efcc32d094e9fb9e1a88e_step_003",
|
||||
"00000844_2c1efcc32d094e9fb9e1a88e_step_004",
|
||||
"00000849_2c1efcc32d094e9fb9e1a88e_step_009",
|
||||
"00000850_2c1efcc32d094e9fb9e1a88e_step_010",
|
||||
"00000851_2c1efcc32d094e9fb9e1a88e_step_011",
|
||||
"00000855_2c1efcc32d094e9fb9e1a88e_step_015",
|
||||
"00000856_2c1efcc32d094e9fb9e1a88e_step_016",
|
||||
"00000859_58ac6a347bbe4cf5ae07a565_step_000",
|
||||
"00000863_029c2575dd474650ba1b8d21_step_003",
|
||||
"00000864_029c2575dd474650ba1b8d21_step_004",
|
||||
"00000866_029c2575dd474650ba1b8d21_step_006",
|
||||
"00000867_029c2575dd474650ba1b8d21_step_007",
|
||||
"00000868_029c2575dd474650ba1b8d21_step_008",
|
||||
"00000870_029c2575dd474650ba1b8d21_step_010",
|
||||
"00000871_029c2575dd474650ba1b8d21_step_011",
|
||||
"00000872_029c2575dd474650ba1b8d21_step_012",
|
||||
"00000873_029c2575dd474650ba1b8d21_step_013",
|
||||
"00000874_029c2575dd474650ba1b8d21_step_014",
|
||||
"00000877_029c2575dd474650ba1b8d21_step_017",
|
||||
"00000879_029c2575dd474650ba1b8d21_step_019",
|
||||
"00000880_029c2575dd474650ba1b8d21_step_020",
|
||||
"00000882_029c2575dd474650ba1b8d21_step_022",
|
||||
"00000883_029c2575dd474650ba1b8d21_step_023",
|
||||
"00000884_029c2575dd474650ba1b8d21_step_024",
|
||||
"00000885_029c2575dd474650ba1b8d21_step_025",
|
||||
"00000886_029c2575dd474650ba1b8d21_step_026",
|
||||
"00000887_029c2575dd474650ba1b8d21_step_027",
|
||||
"00000938_4f351b53bfd84bb4a2f7809c_step_003",
|
||||
"00000970_edacb215ec6949d8af6c8c68_step_000",
|
||||
"00000993_ac4ba032d4b5431ca04f542c_step_000",
|
||||
"00000994_ac4ba032d4b5431ca04f542c_step_001",
|
||||
"00000998_6962bc8aad2c4e27b046d5b4_step_000",
|
||||
"00001000_6962bc8aad2c4e27b046d5b4_step_002",
|
||||
"00001002_6962bc8aad2c4e27b046d5b4_step_004",
|
||||
"00001003_6962bc8aad2c4e27b046d5b4_step_005",
|
||||
"00001004_6962bc8aad2c4e27b046d5b4_step_006",
|
||||
"00001005_6962bc8aad2c4e27b046d5b4_step_007",
|
||||
"00001006_6962bc8aad2c4e27b046d5b4_step_008",
|
||||
"00001009_6962bc8aad2c4e27b046d5b4_step_011",
|
||||
"00001010_6962bc8aad2c4e27b046d5b4_step_012",
|
||||
"00001011_6962bc8aad2c4e27b046d5b4_step_013",
|
||||
"00001013_6962bc8aad2c4e27b046d5b4_step_015",
|
||||
"00001014_6962bc8aad2c4e27b046d5b4_step_016",
|
||||
"00001015_6962bc8aad2c4e27b046d5b4_step_017",
|
||||
"00001034_085828cd621d404a90ccd883_step_000",
|
||||
"00001063_c6fda10dc58448d1951c5cc3_step_000",
|
||||
"00001064_c6fda10dc58448d1951c5cc3_step_001",
|
||||
"00001065_c6fda10dc58448d1951c5cc3_step_002",
|
||||
"00001067_8a2edcf36bcd435092f4c8ae_step_001",
|
||||
"00001069_8a2edcf36bcd435092f4c8ae_step_003",
|
||||
"00001071_8a2edcf36bcd435092f4c8ae_step_005",
|
||||
"00001112_3bd4a008a3a0483d86ccab3f_step_000",
|
||||
"00001177_9f1ce06d5e334fe88e1581aa_step_001",
|
||||
"00001178_515d7e1aeb364a2d88083cf3_step_000",
|
||||
"00001262_157e5348bf6545e282472856_step_000",
|
||||
"00001263_157e5348bf6545e282472856_step_001",
|
||||
"00001305_3f3a389e9d5a4370b952bdce_step_001",
|
||||
"00001310_3f3a389e9d5a4370b952bdce_step_006",
|
||||
"00001358_2e314c6e080b40fcb6a6db96_step_001",
|
||||
"00001360_2e314c6e080b40fcb6a6db96_step_003",
|
||||
"00001368_2e314c6e080b40fcb6a6db96_step_011",
|
||||
"00001369_2e314c6e080b40fcb6a6db96_step_012",
|
||||
"00001371_2e314c6e080b40fcb6a6db96_step_014",
|
||||
"00001373_2e314c6e080b40fcb6a6db96_step_016",
|
||||
"00001375_2e314c6e080b40fcb6a6db96_step_018",
|
||||
"00001377_2e314c6e080b40fcb6a6db96_step_020",
|
||||
"00001398_49414db2dd594dc6a3907cf3_step_001",
|
||||
"00001399_49414db2dd594dc6a3907cf3_step_002",
|
||||
"00001400_49414db2dd594dc6a3907cf3_step_003",
|
||||
"00001401_49414db2dd594dc6a3907cf3_step_004",
|
||||
"00001402_49414db2dd594dc6a3907cf3_step_005",
|
||||
"00001403_49414db2dd594dc6a3907cf3_step_006",
|
||||
"00001404_49414db2dd594dc6a3907cf3_step_007",
|
||||
"00001405_49414db2dd594dc6a3907cf3_step_008",
|
||||
"00001406_49414db2dd594dc6a3907cf3_step_009",
|
||||
"00001407_49414db2dd594dc6a3907cf3_step_010",
|
||||
"00001408_49414db2dd594dc6a3907cf3_step_011",
|
||||
"00001409_49414db2dd594dc6a3907cf3_step_012",
|
||||
"00001410_49414db2dd594dc6a3907cf3_step_013",
|
||||
"00001411_49414db2dd594dc6a3907cf3_step_014",
|
||||
"00001412_49414db2dd594dc6a3907cf3_step_015",
|
||||
"00001413_4642a707b7894949ba816427_step_000",
|
||||
"00001522_7079a898746d41ba8ae13d69_step_003",
|
||||
"00001598_c118664e9f6f4943855293fa_step_000"
|
||||
],
|
||||
"external_fold_previously_observed": true,
|
||||
"epochs": 200,
|
||||
"U": 32000,
|
||||
"twoU": 64000,
|
||||
"delta": 0.001,
|
||||
"bootstrap": {
|
||||
"replicates": 2000,
|
||||
"seed": 2026111408,
|
||||
"unit": "paired evaluation component",
|
||||
"joint_resampling_all_contrasts": true
|
||||
},
|
||||
"endpoints": [
|
||||
"epoch25",
|
||||
"epoch50",
|
||||
"S1_epoch100_reuse",
|
||||
"step50_U",
|
||||
"step50_2U",
|
||||
"step100_U",
|
||||
"step100_2U"
|
||||
],
|
||||
"clock_baselines": [
|
||||
"c25F",
|
||||
"c50F",
|
||||
"c100F"
|
||||
],
|
||||
"domains": [
|
||||
"own_learning",
|
||||
"common_validation",
|
||||
"common_test"
|
||||
],
|
||||
"metrics": [
|
||||
"relative_l2",
|
||||
"last_location",
|
||||
"kendall_tau_b"
|
||||
],
|
||||
"component_counts": {
|
||||
"25": 34,
|
||||
"50": 65,
|
||||
"100": 132
|
||||
},
|
||||
"slope": "(E50_H-E100_H)/log2(n100/n50), H=U and2U",
|
||||
"gain": "E_n_U-E_n_2U for n50,n100",
|
||||
"interaction": "s_2U-s_U; joint component resampling",
|
||||
"classifications": {
|
||||
"data_limited": "s2U low>delta and g100 entireCI in[-delta,delta]",
|
||||
"updates_limited": "s2U entireCI in[-delta,delta] and g100 low>delta",
|
||||
"both": "both low>delta",
|
||||
"local_plateau": "both entireCI in[-delta,delta]",
|
||||
"otherwise": "UNIDENTIFIED_OR_OVERFIT_SIGNS"
|
||||
},
|
||||
"undefined": "Report missing/undefined support; do not call wide/cross-zero CI plateau",
|
||||
"not_training_randomness": true,
|
||||
"no_3U_or_extrapolation": true,
|
||||
"predictions_frozen_before_external_targets": true,
|
||||
"new_exact_postfit_review_required": true,
|
||||
"revision": "Claude1794948643 complete-schedule amendment to e24fff7c",
|
||||
"schedule": "Four separate fresh step-budget fits, own complete warmup3200+cosine endpoint",
|
||||
"update_effect": "joint update budget plus complete LR schedule, not pure continuation",
|
||||
"new_fits": 6,
|
||||
"old_entry_never_started": true
|
||||
}
|
||||
File diff suppressed because it is too large.
Load diff
@@ -0,0 +1,54 @@
|
||||
{
|
||||
"reviewed_commit": "53cca79a47caebaa1ddb8a10e201c1b050dfa258",
|
||||
"contract_sha256": "4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa",
|
||||
"implementation_proof_sha256": "0a073cef46a486947b7c0655b722bfda61776ab0d82bf19ac4abd09122f0103d",
|
||||
"comparison_contract_sha256": "332f2515911c99aadf2486086ddff82bd195d889faa8391296443e7aeacdfee9",
|
||||
"code_sha256": {
|
||||
"/home/frank14f/optane/MeltMind/tools/test_phase3_item1_postfit.py": "7b2187d54a611924e66bc64cfcc7027f421c05a08eb7744889de64a8feba05ba",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_guard.py": "f76a0f052ca474f0d49a7ee042699356c3012679cc323dbcba2836d5d5f0799b",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_gpu_supervisor.py": "689c0f3179e4fcd362e546d90214829a9e4f8739abdbdf74b560763e5a392f15",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_monitoring.py": "442c2a9c75e8c00b30a8f2e91cb52d87c466b19a80aea9128b10c021e61b922e",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_precision_continuation_guard.py": "01235d897c7c03e5ffa109c58b14f5d24d128e8f1ba873e1f76cfb14e02f7a7f",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_gpu_idle_gate.py": "e2b1613e7e6ea398a8d555220a9365a1ad998ab4ef4b2bf8ec7cc4e31f8b4b9f",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_pilot_train.py": "e9cceead51a5cc092aad23409ad98909c84a62e7e10571e750d1e9be1ad3b948",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/network.py": "5137e845d81d1223afc13bf87e03ccba1132a3baa75d5189a8f320b2752be071",
|
||||
"/home/frank14f/optane/MeltMind/tools/freeze_g1_components.py": "d0cee60d95243ac2c59ef3f2244ba818315718c4e3d1895f2f9f064f1f7f5ecd",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_main_memory_guard.py": "e094f5443e11518efe9018215c11718a8e18aa462a977bf43e974b2d0b2aebb1",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h1_worker_v2.py": "1543df85b615bb69b79bacd572074a7c77ae19dab110197601573a7e5805ed20",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_residual_math.py": "ffbd9624dec79bdaabf95244fe87209055908f31a977e73ea4377045d6f8fe62",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_models.py": "53b6fee8b700b9b982b06d29908ce2ce23e0c57512c46dbf74b2aade79e3ac3a",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_data.py": "de9cbcc147677c9144c706acbe52c0f7d505dad17d741d86dcbd04d052f940e2",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_model.py": "fab2470dd0654ca74c3b7d893f35ad1f3ae6c5c08ffe334ef6312cf6ec87892a",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_data.py": "7e03dd7a94e7bb43593277f183c222473d7cb7aaf2f60e53b45ab8d9499fcb09",
|
||||
"/home/frank14f/optane/MeltMind/tools/test_phase2_gate_outcome_postfit.py": "ae3f0849b8cf0abba75d3a7fd7b14d756234a4c094c23e446c5aaa61b75d7c79",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_custody_statistics.py": "188565801f6901d0192f79aec0889e632501e672a4149d76d1835efbb9b45447",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/objectives.py": "5b6527a45f08690055d572335470ce5979e8bb86e1cab00de594f79ae21316b1",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_postfit.py": "b05ee93cfac1f032c84b5a0e33e6d052a4f3b507ad6c717aa6d42456ae4bc512",
|
||||
"/home/frank14f/optane/MeltMind/tools/r4bc_guard.py": "437a756ddf10b125dce450f463e76e28df52724f8c5766e0a989dfa536805628",
|
||||
"/home/frank14f/optane/MeltMind/tools/r4bc_common.py": "27543fff987975ece767167f9771e6621c825035a053805604e4a696794dfcbd",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_supervisor.py": "f563ca3c2665f5baf0c04f296616e4640423a59b815fbb00479af0f29e57a919",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/features.py": "3b34ae61cf8012ebd58b7505a73d7aa43d8ee1890ac54f98d7917b67887a45a4",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_custody_common.py": "588c646101be0d50f505f9a12a1c98361aa3437c8c594eecb7321831f717a741",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_b1213_features.py": "d8e975b0a5362be7ff00fbf554f2b438b54af1b6b95f8993e7c2beae77f2c8f2",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_b12_math.py": "5465b34be5578cc62a7afc67e3cee3cd489ebb75ce6622a3fa320d1b81196c57",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_v2.py": "7bfc81d820a1be7d40b97b2c0b65a616812cb7615b99d80c4ac589788282c291",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_partition.py": "5e7dbf97604b9ea9646113a9847e7232020e5f9248fad388177c32ae2a7117da",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_prepare.py": "1a695477ee6605d75e7e991205e3e6a49a0907a7733491af8d2771c7bc6a5c23",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_features.py": "92bca4d348f31246c2e931949e2ca38103881b1b784a621a42044db56a460026",
|
||||
"/home/frank14f/optane/MeltMind/tools/boundary_distance.py": "a45c1cbce92521a96eb814913b4f465b926c0b3883932ab0b7c37d11db113ed7",
|
||||
"/home/frank14f/optane/MeltMind/tools/audit_samples.py": "a3daf70ec3bcb77bf46d9dc4887d552cf7c422117880a169969780012278cd70",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_features_v2.py": "7dcd5ac218ffdc1ca149edcd87e8658fb3c59e18c0fb9ce47985c7f1357d2cdf",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_statistics.py": "8e9f781cc4975beff245d991a9b886e0d8e7bf6d66c0ff0337d76dccaddda8e1",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_review.py": "915617aeb8292d03c654e637bc34a014c254fd178a773f135d1806c4b41372a8",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_home_freeze_io.py": "bf09204b85ed2a7038acdd202257f618474d1110a9d6d772f105a3ffb334d3ff",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_local_freeze_io.py": "60adf3382577613c5234f5cf30779336b9951b66e90c075cb7cd0adab9cc42dd",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_fixed_io.py": "ea17fdfc7508f3f980e3cdf69534b1eb8ceb8a8e469c32451b2d5e16d0ada575",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_worker.py": "987c26135c9ec3f4411925d16184e091c874a79a3b8935e72e39a3cbdb80ebed",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_statistics.py": "1bb63d4ccf111280dd1a34cec7aeaa28026acc319a27364eb0ff57efee4d1e11",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_common.py": "f203400f33bb345dee46eac7599f24b8938084124d7b6df1b3bc171537611e6e",
|
||||
"/home/frank14f/optane/MeltMind/tools/prepare_phase3_item1_postfit.py": "e17bdf3e123a27c9187915945c17da31f24af593760a71ba91e321e5eb998a06",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_prepare.py": "3a31af0766f628a80f38e0b7286beb11ca3109188eabc4adfdb2ecb0eef3dc27",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit.py": "076a1a07b685a8fe5208954fdc6365cf5176f0efaf785697fdfbdc5614074569",
|
||||
"/home/frank14f/optane/MeltMind/docs/phase3-item1-postfit-2026-11-18.md": "cead5f84314befe8943333cb66cafb243e8cba408b539f70d62b6fed9baa2d1e"
|
||||
}
|
||||
}
|
||||
@@ -0,0 +1,4 @@
|
||||
{
|
||||
"contract_sha256": "4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa",
|
||||
"proof_sha256": "0a073cef46a486947b7c0655b722bfda61776ab0d82bf19ac4abd09122f0103d"
|
||||
}
|
||||
@@ -0,0 +1,117 @@
|
||||
{
|
||||
"commit": "53cca79a47caebaa1ddb8a10e201c1b050dfa258",
|
||||
"files": {
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/contract.json": "4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/comparison-contract.json": "332f2515911c99aadf2486086ddff82bd195d889faa8391296443e7aeacdfee9",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/acceptance.json": "e1c44c93e28b86c28d48a3586c978e8bf2467c7ebdb04e6450cd5c287d40a4d8",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/preparation-audit.json": "8eee4679f178a3e138ff9b6304fe8ed68989f24cb8ecb3f3d4e0a71df5d7084a",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/training-live-snapshot.json": "02d81fec685ef2c070d42cf8055054cd6d76cdbd5926cb7d8d3d8b7ea4693d92",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt1.log": "4a82fe8f640c80de7f93bc86d543dd3f62a2f5209e2b11bb0c34d3b239ded763",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt2.log": "e151d0c3794b979e77f897eb6159b427833a07c73dd889214800894eec62b1b8",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt3.log": "6c452434cfe3d4fbde4ce6a9278b6c49c111c603dd2deabc5a24da33d380269a",
|
||||
"/home/frank14f/optane/MeltMind/tools/test_phase3_item1_postfit.py": "7b2187d54a611924e66bc64cfcc7027f421c05a08eb7744889de64a8feba05ba",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_guard.py": "f76a0f052ca474f0d49a7ee042699356c3012679cc323dbcba2836d5d5f0799b",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_gpu_supervisor.py": "689c0f3179e4fcd362e546d90214829a9e4f8739abdbdf74b560763e5a392f15",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_monitoring.py": "442c2a9c75e8c00b30a8f2e91cb52d87c466b19a80aea9128b10c021e61b922e",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_precision_continuation_guard.py": "01235d897c7c03e5ffa109c58b14f5d24d128e8f1ba873e1f76cfb14e02f7a7f",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_gpu_idle_gate.py": "e2b1613e7e6ea398a8d555220a9365a1ad998ab4ef4b2bf8ec7cc4e31f8b4b9f",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_pilot_train.py": "e9cceead51a5cc092aad23409ad98909c84a62e7e10571e750d1e9be1ad3b948",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/network.py": "5137e845d81d1223afc13bf87e03ccba1132a3baa75d5189a8f320b2752be071",
|
||||
"/home/frank14f/optane/MeltMind/tools/freeze_g1_components.py": "d0cee60d95243ac2c59ef3f2244ba818315718c4e3d1895f2f9f064f1f7f5ecd",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_main_memory_guard.py": "e094f5443e11518efe9018215c11718a8e18aa462a977bf43e974b2d0b2aebb1",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h1_worker_v2.py": "1543df85b615bb69b79bacd572074a7c77ae19dab110197601573a7e5805ed20",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_residual_math.py": "ffbd9624dec79bdaabf95244fe87209055908f31a977e73ea4377045d6f8fe62",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_models.py": "53b6fee8b700b9b982b06d29908ce2ce23e0c57512c46dbf74b2aade79e3ac3a",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_data.py": "de9cbcc147677c9144c706acbe52c0f7d505dad17d741d86dcbd04d052f940e2",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_model.py": "fab2470dd0654ca74c3b7d893f35ad1f3ae6c5c08ffe334ef6312cf6ec87892a",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_data.py": "7e03dd7a94e7bb43593277f183c222473d7cb7aaf2f60e53b45ab8d9499fcb09",
|
||||
"/home/frank14f/optane/MeltMind/tools/test_phase2_gate_outcome_postfit.py": "ae3f0849b8cf0abba75d3a7fd7b14d756234a4c094c23e446c5aaa61b75d7c79",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_custody_statistics.py": "188565801f6901d0192f79aec0889e632501e672a4149d76d1835efbb9b45447",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/objectives.py": "5b6527a45f08690055d572335470ce5979e8bb86e1cab00de594f79ae21316b1",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_postfit.py": "b05ee93cfac1f032c84b5a0e33e6d052a4f3b507ad6c717aa6d42456ae4bc512",
|
||||
"/home/frank14f/optane/MeltMind/tools/r4bc_guard.py": "437a756ddf10b125dce450f463e76e28df52724f8c5766e0a989dfa536805628",
|
||||
"/home/frank14f/optane/MeltMind/tools/r4bc_common.py": "27543fff987975ece767167f9771e6621c825035a053805604e4a696794dfcbd",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_supervisor.py": "f563ca3c2665f5baf0c04f296616e4640423a59b815fbb00479af0f29e57a919",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/features.py": "3b34ae61cf8012ebd58b7505a73d7aa43d8ee1890ac54f98d7917b67887a45a4",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_custody_common.py": "588c646101be0d50f505f9a12a1c98361aa3437c8c594eecb7321831f717a741",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_b1213_features.py": "d8e975b0a5362be7ff00fbf554f2b438b54af1b6b95f8993e7c2beae77f2c8f2",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_b12_math.py": "5465b34be5578cc62a7afc67e3cee3cd489ebb75ce6622a3fa320d1b81196c57",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_v2.py": "7bfc81d820a1be7d40b97b2c0b65a616812cb7615b99d80c4ac589788282c291",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_partition.py": "5e7dbf97604b9ea9646113a9847e7232020e5f9248fad388177c32ae2a7117da",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_prepare.py": "1a695477ee6605d75e7e991205e3e6a49a0907a7733491af8d2771c7bc6a5c23",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_features.py": "92bca4d348f31246c2e931949e2ca38103881b1b784a621a42044db56a460026",
|
||||
"/home/frank14f/optane/MeltMind/tools/boundary_distance.py": "a45c1cbce92521a96eb814913b4f465b926c0b3883932ab0b7c37d11db113ed7",
|
||||
"/home/frank14f/optane/MeltMind/tools/audit_samples.py": "a3daf70ec3bcb77bf46d9dc4887d552cf7c422117880a169969780012278cd70",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_features_v2.py": "7dcd5ac218ffdc1ca149edcd87e8658fb3c59e18c0fb9ce47985c7f1357d2cdf",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_statistics.py": "8e9f781cc4975beff245d991a9b886e0d8e7bf6d66c0ff0337d76dccaddda8e1",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_review.py": "915617aeb8292d03c654e637bc34a014c254fd178a773f135d1806c4b41372a8",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_home_freeze_io.py": "bf09204b85ed2a7038acdd202257f618474d1110a9d6d772f105a3ffb334d3ff",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_local_freeze_io.py": "60adf3382577613c5234f5cf30779336b9951b66e90c075cb7cd0adab9cc42dd",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_fixed_io.py": "ea17fdfc7508f3f980e3cdf69534b1eb8ceb8a8e469c32451b2d5e16d0ada575",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_worker.py": "987c26135c9ec3f4411925d16184e091c874a79a3b8935e72e39a3cbdb80ebed",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_statistics.py": "1bb63d4ccf111280dd1a34cec7aeaa28026acc319a27364eb0ff57efee4d1e11",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_common.py": "f203400f33bb345dee46eac7599f24b8938084124d7b6df1b3bc171537611e6e",
|
||||
"/home/frank14f/optane/MeltMind/tools/prepare_phase3_item1_postfit.py": "e17bdf3e123a27c9187915945c17da31f24af593760a71ba91e321e5eb998a06",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_prepare.py": "3a31af0766f628a80f38e0b7286beb11ca3109188eabc4adfdb2ecb0eef3dc27",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit.py": "076a1a07b685a8fe5208954fdc6365cf5176f0efaf785697fdfbdc5614074569",
|
||||
"/home/frank14f/optane/MeltMind/docs/phase3-item1-postfit-2026-11-18.md": "cead5f84314befe8943333cb66cafb243e8cba408b539f70d62b6fed9baa2d1e"
|
||||
},
|
||||
"git_paths": {
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/contract.json": "home-runs/phase3-item1-postfit-2026-11-18/contract.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/comparison-contract.json": "home-runs/phase3-item1-postfit-2026-11-18/comparison-contract.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/acceptance.json": "home-runs/phase3-item1-postfit-2026-11-18/acceptance.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/preparation-audit.json": "home-runs/phase3-item1-postfit-2026-11-18/preparation-audit.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/training-live-snapshot.json": "home-runs/phase3-item1-postfit-2026-11-18/training-live-snapshot.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt1.log": "home-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt1.log",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt2.log": "home-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt2.log",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt3.log": "home-runs/phase3-item1-postfit-2026-11-18/synthetic-attempt3.log",
|
||||
"/home/frank14f/optane/MeltMind/tools/test_phase3_item1_postfit.py": "tools/test_phase3_item1_postfit.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_guard.py": "tools/phase3_item1_postfit_guard.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_gpu_supervisor.py": "tools/s2_recovery_gpu_supervisor.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_monitoring.py": "tools/s2_recovery_monitoring.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_precision_continuation_guard.py": "tools/s2_precision_continuation_guard.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_gpu_idle_gate.py": "tools/s2_gpu_idle_gate.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_pilot_train.py": "tools/s2_pilot_train.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/network.py": "tools/s2_models/network.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/freeze_g1_components.py": "tools/freeze_g1_components.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_main_memory_guard.py": "tools/s2_main_memory_guard.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h1_worker_v2.py": "tools/phase2_h1_worker_v2.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_residual_math.py": "tools/phase2_residual_math.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_models.py": "tools/gre24_models.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_data.py": "tools/phase3_item1_postfit_data.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_model.py": "tools/phase2_h2_model.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_data.py": "tools/phase2_h2_data.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/test_phase2_gate_outcome_postfit.py": "tools/test_phase2_gate_outcome_postfit.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_custody_statistics.py": "tools/s2_custody_statistics.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/objectives.py": "tools/s2_models/objectives.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_postfit.py": "tools/phase2_gate_outcome_postfit.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/r4bc_guard.py": "tools/r4bc_guard.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/r4bc_common.py": "tools/r4bc_common.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_recovery_supervisor.py": "tools/s2_recovery_supervisor.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_models/features.py": "tools/s2_models/features.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/s2_custody_common.py": "tools/s2_custody_common.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_b1213_features.py": "tools/phase2_b1213_features.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_b12_math.py": "tools/phase2_b12_math.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_v2.py": "tools/o1_v2.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_partition.py": "tools/o1_partition.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/o1_prepare.py": "tools/o1_prepare.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_features.py": "tools/gre24_features.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/boundary_distance.py": "tools/boundary_distance.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/audit_samples.py": "tools/audit_samples.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/gre24_features_v2.py": "tools/gre24_features_v2.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_statistics.py": "tools/phase2_gate_outcome_statistics.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_gate_outcome_review.py": "tools/phase2_gate_outcome_review.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_home_freeze_io.py": "tools/phase2_home_freeze_io.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_local_freeze_io.py": "tools/phase2_local_freeze_io.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_fixed_io.py": "tools/phase2_fixed_io.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_worker.py": "tools/phase3_item1_postfit_worker.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_statistics.py": "tools/phase3_item1_postfit_statistics.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit_common.py": "tools/phase3_item1_postfit_common.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/prepare_phase3_item1_postfit.py": "tools/prepare_phase3_item1_postfit.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase2_h2_prepare.py": "tools/phase2_h2_prepare.py",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_item1_postfit.py": "tools/phase3_item1_postfit.py",
|
||||
"/home/frank14f/optane/MeltMind/docs/phase3-item1-postfit-2026-11-18.md": "docs/phase3-item1-postfit-2026-11-18.md"
|
||||
},
|
||||
"remote_published": false,
|
||||
"local_repository": "/tmp/meltmind-phase2-local-fixed-20261117"
|
||||
}
|
||||
@@ -0,0 +1,108 @@
|
||||
{
|
||||
"status": "PREPARED_NOT_LAUNCHED",
|
||||
"unix": 1794951833.1438994,
|
||||
"metadata_cpu_seconds": 0.775109091,
|
||||
"training_implementation": "73f94bbc77e5ed3156ec85ff4a32ec2c834fdb07",
|
||||
"training_code_pins_verified": 26,
|
||||
"training_processes": [
|
||||
{
|
||||
"pid": 993672,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"/home/frank14f/optane/MeltMind/tools/phase3_data_sufficiency_v2_queue.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993674,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_guard.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"0"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993675,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_guard.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"1"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993676,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_guard.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"2"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993677,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_guard.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"3"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993709,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_worker.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"2"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993719,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_worker.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"1"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993724,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_worker.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"3"
|
||||
]
|
||||
},
|
||||
{
|
||||
"pid": 993725,
|
||||
"command": [
|
||||
"/home/frank14f/optane/MeltMind/artifacts/gre24-reproduction-2026-11-15/environment/venv/bin/python",
|
||||
"tools/phase3_data_sufficiency_v2_worker.py",
|
||||
"/home/frank14f/MeltMind-runs/phase3-data-sufficiency-v2-2026-11-18",
|
||||
"0"
|
||||
]
|
||||
}
|
||||
],
|
||||
"training_complete_exists": false,
|
||||
"future_new_checkpoint_hashes_not_adopted": true,
|
||||
"comparison_byte_identical": true,
|
||||
"S1_prediction_hashes_checked": 121,
|
||||
"expected_prediction_entries": 4582,
|
||||
"model_entries": 3265,
|
||||
"new_model_entries": 3144,
|
||||
"reused_model_entries": 121,
|
||||
"clock_entries": 1317,
|
||||
"unique_parts": 552,
|
||||
"test_parts": 121,
|
||||
"new_fits": 0,
|
||||
"new_predictions": 0,
|
||||
"new_scores": 0,
|
||||
"test_target_values_read": false,
|
||||
"old_score_values_read": false,
|
||||
"training_untouched": true,
|
||||
"new_review_pending": true,
|
||||
"no_waiter_started": true
|
||||
}
|
||||
@@ -0,0 +1,4 @@
|
||||
{
|
||||
"contract_sha256": "4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa",
|
||||
"proof_sha256": "cf10bc6faac01f289b0b8613bcc9add75edf6d91aac02fb20ced2f05a1493bb3"
|
||||
}
|
||||
@@ -0,0 +1,17 @@
|
||||
{
|
||||
"commit": "8fcd30cfb35251fd19da3ca1a021987960601183",
|
||||
"files": {
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/contract.json": "4229906e5d2a6e7febee84b932e2e45b817421f255127cbfb0fe66cc74d6e9aa",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/expected-review-binding.json": "c196b21571ed901e7a8aa3874b34c53fb8a4ef798e6a5694dd804d153fb391b1",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/implementation-git-freeze.json": "0a073cef46a486947b7c0655b722bfda61776ab0d82bf19ac4abd09122f0103d",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/implementation-authorization.json": "614483b04003af2cc06641415412b7ddd8142718dc73ae4f32bc436a92f3ac8b"
|
||||
},
|
||||
"git_paths": {
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/contract.json": "home-runs/phase3-item1-postfit-2026-11-18/contract.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/expected-review-binding.json": "home-runs/phase3-item1-postfit-2026-11-18/expected-review-binding.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/implementation-git-freeze.json": "home-runs/phase3-item1-postfit-2026-11-18/implementation-git-freeze.json",
|
||||
"/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18/implementation-authorization.json": "home-runs/phase3-item1-postfit-2026-11-18/implementation-authorization.json"
|
||||
},
|
||||
"remote_published": false,
|
||||
"local_repository": "/tmp/meltmind-phase2-local-fixed-20261117"
|
||||
}
|
||||
@@ -0,0 +1,23 @@
|
||||
test_S1_checkpoint_selection (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_actual_count_denominator (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_budget_fixed_remaining (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_budget_nonfinite (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_checkpoint_identity (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_component_missing_not_silently_promoted (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_data_boundary (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_every_arm_expected_domain (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_failure_peaks_retained (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_four_classifications (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_incomplete_not_classified (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_joint_draw_cancellation (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_loader_denies_wrong_role_before_IO (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_loader_no_target_branch (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_prediction_no_label_calls (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_query_order_mapping (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_training_pending_refuses_before_adoption (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_wide_not_plateau (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
|
||||
----------------------------------------------------------------------
|
||||
Ran 18 tests in 0.056s
|
||||
|
||||
OK
|
||||
@@ -0,0 +1,32 @@
|
||||
test_S1_checkpoint_selection (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_actual_count_denominator (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_budget_fixed_remaining (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_budget_nonfinite (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_checkpoint_identity (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_component_missing_not_silently_promoted (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_data_boundary (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_every_arm_expected_domain (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_failure_peaks_retained (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_four_classifications (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_incomplete_not_classified (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_joint_draw_cancellation (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_loader_denies_wrong_role_before_IO (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_loader_no_target_branch (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_prediction_no_label_calls (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_query_order_mapping (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_score_rejects_missing_PASS_before_targets (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_score_rejects_missing_prediction_freeze_before_targets (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_six_terminal_identity_and_once_ledger (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_training_pending_refuses_before_adoption (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_wide_not_plateau (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_actual_frozen_pass (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_changed_comparison_rejected (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_changed_source_rejected (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_fail_missing_and_unrelated_receipts (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_review_omitted_from_freeze (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_wrong_exact_identities (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
|
||||
----------------------------------------------------------------------
|
||||
Ran 27 tests in 0.636s
|
||||
|
||||
OK
|
||||
@@ -0,0 +1,32 @@
|
||||
test_S1_checkpoint_selection (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_actual_count_denominator (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_budget_fixed_remaining (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_budget_nonfinite (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_checkpoint_identity (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_component_missing_not_silently_promoted (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_data_boundary (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_every_arm_expected_domain (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_failure_peaks_retained (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_four_classifications (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_incomplete_not_classified (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_joint_draw_cancellation (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_loader_denies_wrong_role_before_IO (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_loader_no_target_branch (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_prediction_no_label_calls (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_query_order_mapping (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_score_rejects_missing_PASS_before_targets (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_score_rejects_missing_prediction_freeze_before_targets (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_six_terminal_identity_and_once_ledger (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_training_pending_refuses_before_adoption (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_wide_not_plateau (tools.test_phase3_item1_postfit.PostfitTests) ... ok
|
||||
test_actual_frozen_pass (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_changed_comparison_rejected (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_changed_source_rejected (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_fail_missing_and_unrelated_receipts (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_review_omitted_from_freeze (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
test_wrong_exact_identities (test_phase2_gate_outcome_postfit.SemanticReview) ... ok
|
||||
|
||||
----------------------------------------------------------------------
|
||||
Ran 27 tests in 1.317s
|
||||
|
||||
OK
|
||||
@@ -0,0 +1,31 @@
|
||||
{
|
||||
"stage": "ITEM1_TRAINING",
|
||||
"unix": 1794951832.5333512,
|
||||
"active": [
|
||||
{
|
||||
"id": 0,
|
||||
"guard_pid": 993674
|
||||
},
|
||||
{
|
||||
"id": 1,
|
||||
"guard_pid": 993675
|
||||
},
|
||||
{
|
||||
"id": 2,
|
||||
"guard_pid": 993676
|
||||
},
|
||||
{
|
||||
"id": 3,
|
||||
"guard_pid": 993677
|
||||
}
|
||||
],
|
||||
"terminal": [],
|
||||
"pending": [
|
||||
4,
|
||||
5
|
||||
],
|
||||
"charged_training_gpu_seconds": 10510.522380944109,
|
||||
"postfit_gpu_reserve": 7200,
|
||||
"item_cap": 120000,
|
||||
"phase3_cap": 612000
|
||||
}
|
||||
@@ -0,0 +1,39 @@
|
||||
"""One item1 continuation; immutable model and prediction barriers before custody."""
|
||||
import fcntl,json,os,resource,sys,time
|
||||
from pathlib import Path
|
||||
import psutil
|
||||
from phase3_item1_postfit_common import config,read,terminal_inputs,manifest_identity,require_budget
|
||||
from phase2_gate_outcome_review import validate_review
|
||||
from phase2_home_freeze_io import freeze,verify
|
||||
from phase2_h1_worker_v2 import save,sha,checked
|
||||
from r4bc_guard import supervise
|
||||
|
||||
def run(root):
|
||||
gpu_clock=time.monotonic();start=time.process_time();u0=resource.getrusage(resource.RUSAGE_CHILDREN);c,tc,pilot=config(root);validate_review(root);adopted=terminal_inputs(c,tc);lock=(root/'run.lock').open('a');fcntl.flock(lock,fcntl.LOCK_EX|fcntl.LOCK_NB)
|
||||
if (root/'started.json').exists():raise FileExistsError('One item1 postfit attempt, no retry')
|
||||
for p in psutil.process_iter(['pid','cmdline']):
|
||||
if p.pid!=os.getpid() and any(Path(x).name in {'phase3_item1_postfit.py','phase3_item1_postfit_worker.py'} for x in p.info['cmdline'] or []):raise RuntimeError('Existing postfit process')
|
||||
os.sched_setaffinity(0,c['coordinator_cpus']);os.environ['PYTORCH_CUDA_ALLOC_CONF']='expandable_segments:True';gpu=0.;charged=0.;records=[];prior=adopted['training_gpu_seconds_once'];prediction_start=None
|
||||
def cpu():
|
||||
u=resource.getrusage(resource.RUSAGE_CHILDREN);return 180+time.process_time()-start+max(charged,u.ru_utime+u.ru_stime-u0.ru_utime-u0.ru_stime)
|
||||
save(root/'started.json',dict(pid=os.getpid(),unix=time.time(),training_gpu_seconds_once=prior,postfit_gpu_cap=7200,item_gpu_cap=120000,review_sha256=sha(root/'review.json'),cpu_start_counters=dict(self_process_time=start,children_user=u0.ru_utime,children_system=u0.ru_stime),cpu_setup_allowance=180,new_fits=0,new_decodes=0))
|
||||
try:
|
||||
require_budget(prior,gpu,cpu(),7);save(root/'models-adopted.json',adopted);train=Path(c['training_root']);freeze(root,'models',[root/'contract.json',root/'models-adopted.json',train/'complete.json',train/'models-terminal.json',train/'models-git-freeze.json',train/'terminal-git-freeze.json',*[Path(m['checkpoint']) for m in adopted['models']],*[Path(m['standardization']) for m in adopted['models']]])
|
||||
entries=[]
|
||||
from phase3_item1_postfit_guard import run_prediction
|
||||
for index in range(7):
|
||||
gpu=max(gpu,60+time.monotonic()-gpu_clock);require_budget(prior,gpu,cpu(),7-index);(root/'jobs'/str(index)).mkdir(parents=True,exist_ok=False);prediction_start=time.monotonic();g=run_prediction(root,c,pilot,index,prior+gpu,300);observed=60+time.monotonic()-prediction_start;prediction_start=None;cost=max(observed,g['gpu_seconds_charged']);gpu+=cost;charged+=g['worker']['cpu_seconds'] if g['worker'] else 0.;records.append(dict(mode='predict',index=index,charged_gpu_seconds=cost,guard=g));save(root/'resource-progress.json',dict(records=records,prediction_gpu_seconds=gpu,postfit_cpu_seconds=cpu(),cumulative_item_gpu_seconds=prior+gpu))
|
||||
if g['status']!='PASS' or cost>900 or gpu>7200 or prior+gpu>120000:raise RuntimeError('Prediction/resource guard failure')
|
||||
m=read(root/'jobs'/str(index)/'prediction-records/manifest.json');entries.extend(m['entries'])
|
||||
manifest_identity(entries,c,tc,pilot);save(root/'predictions-manifest.json',dict(entries=entries,S1_test_predictions_reused=121,new_fits=0,targets_read=False));freeze(root,'predictions',[root/'contract.json',root/'models-adopted.json',root/'predictions-manifest.json',root/'resource-progress.json',*[Path(p) for p in c['code_sha256']]])
|
||||
gpu=max(gpu,60+time.monotonic()-gpu_clock);require_budget(prior,gpu,cpu(),0);validate_review(root);verify(root,'models');verify(root,'predictions')
|
||||
g=supervise([sys.executable,'tools/phase3_item1_postfit_worker.py','score',str(root)],root/'score-guard',dict(cpu_seconds=900,wall_seconds=1200,rss_bytes=16*2**30,disk_bytes=20*2**30));charged+=g['cpu_seconds'];records.append(dict(mode='score',guard=g))
|
||||
if g['status']!='PASS' or cpu()+120>3600:raise RuntimeError('Score/resource failure')
|
||||
save(root/'complete.json',dict(status='COMPLETE',records=records,prediction_gpu_seconds=gpu,postfit_cpu_seconds=cpu(),training_gpu_seconds_once=prior,cumulative_item_gpu_seconds=prior+gpu,item_gpu_cap=120000,new_fits=0,new_decodes=0,future_phase3_started=False,no_retry=True));freeze(root,'terminal',[root/'contract.json',root/'complete.json',root/'score/result.json',root/'review.json'])
|
||||
final=cpu();save(root/'post-terminal-cost.json',dict(postfit_cpu_seconds=final,prediction_gpu_seconds=gpu,cumulative_item_gpu_seconds=prior+gpu,within_cap=final<=3600 and gpu<=7200 and prior+gpu<=120000,cpu_start_counters=read(root/'started.json')['cpu_start_counters']))
|
||||
if final>3600:raise RuntimeError('Final CPU cap after freeze')
|
||||
except BaseException as exc:
|
||||
if prediction_start is not None:gpu+=60+time.monotonic()-prediction_start
|
||||
if (root/'complete.json').exists():save(root/'pre-failure-complete.json',read(root/'complete.json'))
|
||||
save(root/'complete.json',dict(status='STOPPED',reason=type(exc).__name__+': '+str(exc),records=records,postfit_cpu_seconds=cpu(),prediction_gpu_seconds=gpu,training_gpu_seconds_once=prior,cumulative_item_gpu_seconds=prior+gpu,new_fits=0,no_retry=True));raise
|
||||
if __name__=='__main__':run(Path(sys.argv[1]))
|
||||
@@ -0,0 +1,99 @@
|
||||
"""One reviewed item1 postfit; runtime checkpoint/prediction custody barriers."""
|
||||
import json,math
|
||||
from pathlib import Path
|
||||
from phase2_h1_worker_v2 import checked,sha
|
||||
from phase2_home_freeze_io import verify
|
||||
MODEL_ARMS=('epoch25','epoch50','step50_U','step50_2U','step100_U','step100_2U','S1_epoch100_reuse')
|
||||
DOMAIN={'learning':'own_learning','validation':'common_validation','test':'common_test'}
|
||||
def read(p):return json.loads(Path(p).read_text())
|
||||
def config(root):
|
||||
c=read(root/'contract.json');verify(root,'implementation')
|
||||
if c['approval']!='1794950613.195499' or c['comparison_contract_sha256']!='332f2515911c99aadf2486086ddff82bd195d889faa8391296443e7aeacdfee9' or c['postfit_gpu_cap']!=7200 or c['postfit_cpu_cap']!=3600 or c['item_gpu_cap']!=120000:raise PermissionError('Frozen item1 postfit scope')
|
||||
for p,h in {**c['source_sha256'],**c['code_sha256']}.items():checked(p,h)
|
||||
tc=read(c['training_contract']);pilot=read(tc['pilot_contract'])
|
||||
verify(Path(c['training_root']),'implementation')
|
||||
for p,h in tc['code_sha256'].items():checked(p,h)
|
||||
if tc['new_fits']!=6 or tc['item_gpu_cap']!=120000 or tc['postfit_gpu_reserve']!=7200:raise ValueError('Training scope mismatch')
|
||||
if [j['arm'] for j in c['jobs']]!=list(MODEL_ARMS):raise ValueError('Fixed seven endpoint identities')
|
||||
return c,tc,pilot
|
||||
|
||||
def expected_checkpoint(job,meta,training_sha):
|
||||
if meta.get('fold')!=0 or meta.get('seed')!=0 or meta.get('optimizer_steps')!=job['target_steps'] or meta.get('contract_sha256')!=training_sha:raise ValueError('Checkpoint fold/seed/step/source identity')
|
||||
if job['id']<6:
|
||||
if meta.get('pool')!=job['pool'] or meta.get('mode')!=job['mode'] or meta.get('role')!=job['tag']:raise ValueError('New checkpoint cell identity')
|
||||
elif meta.get('completed_epochs')!=200 or meta.get('role')!='last_epoch_only':raise ValueError('S1 endpoint identity')
|
||||
|
||||
def terminal_identity(done,mm):
|
||||
if done['status']!='TRAIN_COMPLETE_POSTFIT_PENDING' or len(done['entries'])!=6 or any(e['status']!='PASS' or e['returncode']!=0 for e in done['entries']):raise ValueError('All six terminal PASS required')
|
||||
if {e['id'] for e in done['entries']}!=set(range(6)) or len(mm['models'])!=6 or {m['job'] for m in mm['models']}!=set(range(6)):raise ValueError('Six distinct model jobs')
|
||||
if done['test_targets_read'] or mm['test_prediction_started'] or mm['new_fits']!=6 or mm['new_checkpoints']!=6:raise ValueError('Training custody identity')
|
||||
costs=[e['gpu_seconds_charged'] for e in done['entries']];prior=done['training_gpu_seconds']
|
||||
if any(not math.isfinite(x) or x<0 for x in costs+[prior]) or prior>112800 or abs(prior-sum(costs))>1e-8:raise ValueError('Settled training ledger')
|
||||
return prior
|
||||
|
||||
def terminal_inputs(c,tc):
|
||||
import psutil
|
||||
train=Path(c['training_root'])
|
||||
if not (train/'complete.json').exists():raise RuntimeError('WAIT_EXISTING_SIX_TRAINING_TERMINALS')
|
||||
verify(train,'terminal');modelsproof=verify(train,'models');done=read(train/'complete.json');mm=read(train/'models-terminal.json')
|
||||
if (train/'global-stop.json').exists():raise ValueError('Training global stop')
|
||||
prior=terminal_identity(done,mm)
|
||||
for p in psutil.process_iter(['pid','cmdline']):
|
||||
if any(Path(a).name in {'phase3_data_sufficiency_v2_queue.py','phase3_data_sufficiency_v2_guard.py','phase3_data_sufficiency_v2_worker.py'} for a in p.info['cmdline'] or []):raise RuntimeError('Training processes still present')
|
||||
by={m['job']:m for m in mm['models']};adopted=[]
|
||||
for j in c['jobs']:
|
||||
if j['id']<6:
|
||||
m=by[j['id']];w=train/'jobs'/str(j['id'])/'worker';entry=next(e for e in done['entries'] if e['id']==j['id']);g=read(checked(w.parent/'complete.json',entry['guard_complete_sha256']));r=read(checked(w/'result.json',g['worker_result_sha256']))
|
||||
if g['status']!='PASS' or r['status']!='COMPLETE' or r['job']!=j['id'] or r['pool']!=j['pool'] or r['mode']!=j['mode'] or r['optimizer_steps']!=j['target_steps'] or m['tag']!=j['tag'] or m['steps']!=j['target_steps'] or m['pool']!=j['pool'] or m['mode']!=j['mode'] or r['test_targets_read'] or r['test_prediction_started']:raise ValueError('Training worker/manifest identity')
|
||||
if len(r['checkpoints'])!=1 or any(r['checkpoints'][0][k]!=m[k] for k in ['tag','steps','path','sha256','standardization','standardization_sha256']) or r['standardization_sha256']!=m['standardization_sha256']:raise ValueError('Worker checkpoint manifest mismatch')
|
||||
if m['path']!=str(w/(j['tag']+'.pt')) or m['standardization']!=str(w/'standardization.json'):raise ValueError('Checkpoint path identity')
|
||||
checked(m['path'],m['sha256']);stats=read(checked(m['standardization'],m['standardization_sha256']));record=dict(**j,checkpoint=m['path'],checkpoint_sha256=m['sha256'],standardization=m['standardization'],standardization_sha256=m['standardization_sha256'],training_contract_sha256=sha(train/'contract.json'))
|
||||
else:
|
||||
s=tc['S1_reuse'];checked(s['checkpoint'],s['checkpoint_sha256']);stats=read(checked(s['standardization'],s['standardization_sha256']));record=dict(**j,checkpoint=s['checkpoint'],checkpoint_sha256=s['checkpoint_sha256'],standardization=s['standardization'],standardization_sha256=s['standardization_sha256'],training_contract_sha256=c['S1_training_contract_sha256'])
|
||||
if stats['parts']!=tc['pools'][str(j['pool'])]['learning']:raise ValueError('Pool standardization role mismatch')
|
||||
adopted.append(record)
|
||||
return dict(models=adopted,training_gpu_seconds_once=prior,training_terminal_sha256=sha(train/'complete.json'),training_models_sha256=sha(train/'models-terminal.json'),training_models_commit=modelsproof['commit'],new_fits=0,old_S1_GPU_not_recharged=True)
|
||||
|
||||
def adopted_inputs(root,c,tc):
|
||||
verify(root,'models');a=read(root/'models-adopted.json')
|
||||
if len(a['models'])!=7 or [m['arm'] for m in a['models']]!=list(MODEL_ARMS):raise ValueError('Seven adopted endpoints')
|
||||
for m in a['models']:
|
||||
checked(m['checkpoint'],m['checkpoint_sha256']);s=read(checked(m['standardization'],m['standardization_sha256']))
|
||||
if s['parts']!=tc['pools'][str(m['pool'])]['learning']:raise ValueError('Adopted standardization')
|
||||
return a
|
||||
|
||||
def expected_entries(c,tc):
|
||||
out={}
|
||||
for j in c['jobs']:
|
||||
f=tc['pools'][str(j['pool'])]
|
||||
for role,domain in DOMAIN.items():
|
||||
for case in f[role]:out[(j['arm'],case)]=dict(pool=j['pool'],domain=domain)
|
||||
for pool in [25,50,100]:
|
||||
f=tc['pools'][str(pool)]
|
||||
for role,domain in DOMAIN.items():
|
||||
for case in f[role]:out[('c'+str(pool)+'F',case)]=dict(pool=pool,domain=domain)
|
||||
return out
|
||||
|
||||
def manifest_identity(entries,c,tc,pilot):
|
||||
expected=expected_entries(c,tc);keys=[(e['arm'],e['case']) for e in entries]
|
||||
if len(set(keys))!=len(keys) or set(keys)!=set(expected):raise ValueError('Full model/clock/domain manifest required')
|
||||
old={e['case']:e for e in read(tc['S1_reuse']['test_manifest'])['entries']}
|
||||
for e in entries:
|
||||
ex=expected[(e['arm'],e['case'])];row=c['rows'][e['case']]
|
||||
if e['pool']!=ex['pool'] or e['domain']!=ex['domain'] or e['component_id']!=row['component_id'] or e['nodes']!=pilot['parts'][e['case']]['nodes']:raise ValueError('Prediction domain identity')
|
||||
if e['arm']=='S1_epoch100_reuse' and e['domain']=='common_test':
|
||||
o=old[e['case']]
|
||||
if any(e[k]!=o[k] for k in ['path','sha256','nodes','query_nodes','negative_full_predictions','negative_query_predictions']) or e['origin']!='S1_HASH_REUSE':raise ValueError('Exact S1 test reuse')
|
||||
else:
|
||||
p=Path(c['postfit_root'])/'predictions'/e['arm']/(e['case']+'.npz')
|
||||
if e['path']!=str(p) or e['origin']!='ITEM1_FROZEN_ENDPOINT_OR_CLOCK':raise ValueError('New prediction path/origin')
|
||||
for name,n in [('negative_full_predictions',e['nodes']),('negative_query_predictions',e['query_nodes'])]:
|
||||
if type(e[name]) is not int or not 0<=e[name]<=n:raise ValueError('Unclipped count range')
|
||||
checked(e['path'],e['sha256'])
|
||||
return expected
|
||||
|
||||
def require_budget(prior,gpu,cpu,remaining):
|
||||
if any(not math.isfinite(v) or v<0 for v in [prior,gpu,cpu]):raise ValueError('Nonfinite budget ledger')
|
||||
if remaining not in range(8):raise ValueError('Prediction job count')
|
||||
if gpu+900*remaining>7200 or prior+gpu+900*remaining>120000:raise RuntimeError('GPU reserve exhausted')
|
||||
if cpu+300*remaining+900+120>3600:raise RuntimeError('CPU reserve exhausted')
|
||||
@@ -0,0 +1,28 @@
|
||||
"""Original H2 label-free preprocessing; explicit evaluation role and pool; no target branch."""
|
||||
import numpy as np
|
||||
import torch
|
||||
from scipy.sparse import csr_matrix
|
||||
from scipy.sparse.csgraph import connected_components
|
||||
from phase2_h1_worker_v2 import checked
|
||||
from phase2_h2_data import cached
|
||||
from phase2_h2_model import edge_attributes
|
||||
|
||||
def load_evaluation_part(c,case,stats,role):
|
||||
if role not in ('learning','validation','test') or case not in c['folds'][0][role]:raise PermissionError('Frozen evaluation role/pool required')
|
||||
r=c['parts'][case];g=c['geometry'][case]
|
||||
with np.load(checked(g['path'],g['sha256']),allow_pickle=False) as z:
|
||||
points=z['points'].copy();raw=z['raw_ids'].copy();mass=z['mass'].copy()
|
||||
indptr=z['indptr'].copy();sender=z['indices'].astype(np.int64);old_rg=float(z['rg'])
|
||||
n=len(raw);receiver=np.repeat(np.arange(n,dtype=np.int64),np.diff(indptr))
|
||||
if not np.array_equal(raw,cached(c,case,'raw_ids')) or not np.array_equal(mass,cached(c,case,'mass')):raise ValueError('Frozen geometry/node mass identity')
|
||||
if n!=r['nodes'] or len(sender)!=g['directed_edges']:raise ValueError('Frozen graph counts')
|
||||
a,rg=edge_attributes(points,mass,receiver,sender)
|
||||
if not np.isclose(rg,old_rg,rtol=1e-10,atol=1e-12):raise ValueError('Original mass Rg differs')
|
||||
graph=csr_matrix((np.ones(len(sender)),sender,indptr),shape=(n,n));count,comp=connected_components(graph,directed=False)
|
||||
x=cached(c,case,'features');F=cached(c,case,'Vf')
|
||||
if F.dtype!=np.float64 or F.shape!=(n,) or not np.isfinite(F).all():raise ValueError('Original F must be FP64')
|
||||
x=((np.asarray(x,np.float64)-np.asarray(stats['mean']))/np.asarray(stats['scale'])).astype(np.float32)
|
||||
if not np.isfinite(x).all():raise ValueError('Standardized feature overflow')
|
||||
result=dict(x=torch.from_numpy(x),mass=torch.from_numpy(mass.astype(np.float32)),receiver=torch.from_numpy(receiver),sender=torch.from_numpy(sender),attributes=torch.from_numpy(a),component_ids=torch.from_numpy(comp.astype(np.int64)),
|
||||
F=F,raw_ids=raw,physical_components=int(count),nodes=n,directed_edges=len(sender),Rg=rg,t_set=r['t_set'])
|
||||
return result
|
||||
@@ -0,0 +1,74 @@
|
||||
"""Item1 one-endpoint900 GPU-second prediction, retained failure peaks, no retries."""
|
||||
import json,os,signal,sys,threading,time
|
||||
import psutil
|
||||
from phase2_h1_worker_v2 import save
|
||||
from s2_main_memory_guard import PeakRecords
|
||||
from s2_precision_continuation_guard import device_usage
|
||||
from s2_recovery_gpu_supervisor import supervise,tree_usage,stop_group,disk_usage
|
||||
|
||||
def final_reasons(reasons,receipt,result,tracker,gpu_seconds,cpu_seconds,cpu_cap,expected_parts):
|
||||
out=list(reasons)
|
||||
def add(s):
|
||||
if s not in out:out.append(s)
|
||||
if not receipt or receipt['status']!='PASS':add('WORKER_NOT_PASS')
|
||||
if not result or result.get('status')!='COMPLETE' or result.get('parts')!=expected_parts:add('INCOMPLETE_FOLD')
|
||||
if not all(k in tracker.witnesses for k in ['peak_allocated_bytes','peak_reserved_bytes']):add('MISSING_PEAKS')
|
||||
if any(v>15*2**30 for v in tracker.maxima.values()):add('RECORDED_VRAM_LIMIT')
|
||||
if gpu_seconds>900 or cpu_seconds>cpu_cap:add('FINAL_RESOURCE_LIMIT')
|
||||
return out
|
||||
|
||||
def run_prediction(queue,c,pilot,fold,prior_gpu,cpu_cap):
|
||||
root=queue/'jobs'/str(fold);gpu_uuid=c['gpu_uuid'];expected_parts=c['jobs'][fold]['expected_parts']
|
||||
start=time.monotonic();cpu=time.process_time();requested=[];reasons=[];errors=[];samples=[];done=threading.Event();tracker=PeakRecords();receipt=None
|
||||
def elapsed():return 60+time.monotonic()-start
|
||||
def reason(x):
|
||||
if x not in reasons:reasons.append(x)
|
||||
command=[sys.executable,'tools/phase3_item1_postfit_worker.py','predict',str(queue),str(fold)]
|
||||
def kill():
|
||||
for p in psutil.Process().children(recursive=True):
|
||||
try:
|
||||
if p.cmdline()==command:stop_group(p.pid)
|
||||
except psutil.NoSuchProcess:pass
|
||||
def scan():
|
||||
if tracker.scan(root/'prediction-records'):reason('RECORDED_VRAM_LIMIT')
|
||||
def watch():
|
||||
while not done.is_set():
|
||||
try:
|
||||
scan();mem,util=device_usage(gpu_uuid);_,rss,_=tree_usage(os.getpid())
|
||||
if mem>15*2**30:reason('DEVICE_VRAM_LIMIT')
|
||||
if rss>16*2**30:reason('TREE_RSS_LIMIT')
|
||||
if disk_usage(queue)>20*2**30:reason('DISK_LIMIT')
|
||||
if elapsed()>=895:reason('PREDICTION_TIME_LIMIT')
|
||||
for x in requested:reason(x)
|
||||
save(root/'prediction-live.json',dict(gpu_seconds_charged=elapsed(),device_bytes=mem,device_utilization=util,rss_bytes=rss,peaks=tracker.maxima,reasons=reasons,unix=time.time()))
|
||||
except Exception as exc:reason('MONITOR_ERROR');errors.append(str(exc))
|
||||
if reasons:kill()
|
||||
done.wait(.5)
|
||||
old={s:signal.signal(s,lambda signum,frame:requested.append('COORDINATOR_SIGNAL_'+str(signum))) for s in [signal.SIGTERM,signal.SIGINT]};thread=None
|
||||
try:
|
||||
for i in range(3):
|
||||
mem,util=device_usage(gpu_uuid);samples.append(dict(unix=time.time(),bytes=mem,utilization=util))
|
||||
if mem>128*2**20 or util!=0:raise RuntimeError('SELECTED_GPU_BUSY_NO_RETRY')
|
||||
if i<2:time.sleep(2)
|
||||
save(root/'prediction-idle-polls.json',samples)
|
||||
if requested:raise RuntimeError('Signal received during setup')
|
||||
thread=threading.Thread(target=watch,daemon=True);thread.start()
|
||||
remaining=900-elapsed()-10
|
||||
if remaining<=0:raise RuntimeError('Setup exhausted prediction budget')
|
||||
receipt=supervise(command,root/'prediction-guard','predict',dict(total_cpu_seconds=cpu_cap-60,job_cpu_seconds=cpu_cap-60,job_wall_seconds=remaining,tree_rss_bytes=16*2**30,disk_bytes=20*2**30,max_processes=1),poll=.2,gpu_uuid=gpu_uuid)
|
||||
except BaseException as exc:reason(type(exc).__name__+': '+str(exc));kill()
|
||||
finally:
|
||||
done.set()
|
||||
if thread:thread.join(timeout=4)
|
||||
if thread and thread.is_alive():reason('MONITOR_DID_NOT_EXIT')
|
||||
try:scan()
|
||||
except Exception as exc:reason('FINAL_SCAN_ERROR');errors.append(str(exc))
|
||||
for s,h in old.items():signal.signal(s,h)
|
||||
result=None;p=root/'prediction-records/result.json'
|
||||
try:
|
||||
if p.exists():result=json.loads(p.read_text())
|
||||
except Exception as exc:reason('FINAL_RESULT_READ_ERROR');errors.append(str(exc))
|
||||
totalcpu=60+(receipt['cpu_seconds'] if receipt else 0)+time.process_time()-cpu
|
||||
duration=elapsed();reasons=final_reasons(reasons,receipt,result,tracker,duration,totalcpu,cpu_cap,expected_parts)
|
||||
out=dict(status='PASS' if not reasons else 'STOPPED',reasons=reasons,monitor_errors=errors,worker=receipt,recorded_memory=tracker.as_dict(),gpu_seconds_charged=duration,cpu_seconds_charged=totalcpu,cumulative_item1_gpu_seconds=prior_gpu+duration,no_retry=True)
|
||||
save(root/'prediction-guard.json',out);return out
|
||||
@@ -0,0 +1,51 @@
|
||||
"""Frozen item1 contrasts on stored part metrics; common component bootstrap draws."""
|
||||
import math
|
||||
import numpy as np
|
||||
ARMS=('epoch25','epoch50','S1_epoch100_reuse','step50_U','step50_2U','step100_U','step100_2U','c25F','c50F','c100F')
|
||||
METRICS=('relative_l2','last_location','kendall_tau_b')
|
||||
|
||||
def classify(s,g,delta=.001,complete=True):
|
||||
if not complete or any(v is None or not math.isfinite(v) for x in [s,g] for v in [x.get('low'),x.get('high')]):return 'UNIDENTIFIED_OR_OVERFIT_SIGNS'
|
||||
eq=lambda x:-delta<=x['low']<=x['high']<=delta
|
||||
positive=lambda x:x['low']>delta
|
||||
if positive(s) and eq(g):return 'DATA_BENEFIT_AT_TESTED_RANGE'
|
||||
if eq(s) and positive(g):return 'UPDATE_BUDGET_AND_SCHEDULE_BENEFIT'
|
||||
if positive(s) and positive(g):return 'BOTH_BENEFITS'
|
||||
if eq(s) and eq(g):return 'LOCAL_PLATEAU_ONLY'
|
||||
return 'UNIDENTIFIED_OR_OVERFIT_SIGNS'
|
||||
|
||||
def contrasts(a,counts):
|
||||
den=math.log2(counts['100']/counts['50']);den0=math.log2(counts['50']/counts['25'])
|
||||
if den<=0 or den0<=0:raise ValueError('Actual nested component counts')
|
||||
sU=(a['step50_U']-a['step100_U'])/den;s2=(a['step50_2U']-a['step100_2U'])/den
|
||||
out=dict(s_U=sU,s_2U=s2,g_50=a['step50_U']-a['step50_2U'],g_100=a['step100_U']-a['step100_2U'],s_2U_minus_s_U=s2-sU,epoch_s_25to50=(a['epoch25']-a['epoch50'])/den0,epoch_s_50to100=(a['epoch50']-a['S1_epoch100_reuse'])/den,clock_s_25to50=(a['c25F']-a['c50F'])/den0,clock_s_50to100=(a['c50F']-a['c100F'])/den)
|
||||
for arm,pool in [('epoch25',25),('epoch50',50),('S1_epoch100_reuse',100),('step50_U',50),('step50_2U',50),('step100_U',100),('step100_2U',100)]:out[arm+'_minus_clock']=a[arm]-a['c'+str(pool)+'F']
|
||||
return out
|
||||
|
||||
def summarize(rows,counts,replicates=2000,seed=2026111408):
|
||||
out={}
|
||||
for domain in ['own_learning','common_validation','common_test']:
|
||||
domainrows=[r for r in rows if r['domain']==domain];result={}
|
||||
for metric in METRICS:
|
||||
values={};excluded={};parts={}
|
||||
for arm in ARMS:
|
||||
rs=[r for r in domainrows if r['arm']==arm];parts[arm]=len(rs);groups={}
|
||||
for r in rs:groups.setdefault(r['component_id'],[]).append(r)
|
||||
values[arm]={};excluded[arm]={}
|
||||
for g,gr in sorted(groups.items()):
|
||||
x=[r['metrics'][metric]['value'] for r in gr]
|
||||
if any(v is None or not math.isfinite(v) for v in x):excluded[arm][g]=[dict(case=r['case'],reason=r['metrics'][metric].get('reason')) for r in gr if r['metrics'][metric]['value'] is None or not math.isfinite(r['metrics'][metric]['value'])]
|
||||
else:values[arm][g]=float(np.mean(x))
|
||||
result[metric]=dict(means={a:float(np.mean(list(v.values()))) if v else None for a,v in values.items()},parts=parts,components={a:len(v) for a,v in values.items()},excluded=excluded)
|
||||
if domain=='own_learning':
|
||||
result[metric]['contrasts']='Not computed across different learning domains';continue
|
||||
ids=sorted(set.intersection(*(set(values[a]) for a in ARMS)));expected=len({r['component_id'] for r in domainrows});summary={}
|
||||
if ids:
|
||||
arrays={a:np.array([values[a][i] for i in ids]) for a in ARMS};rng=np.random.Generator(np.random.PCG64(np.random.SeedSequence(seed)));draw=rng.integers(0,len(ids),size=(replicates,len(ids)));point=contrasts({a:x.mean() for a,x in arrays.items()},counts);boot=contrasts({a:x[draw].mean(1) for a,x in arrays.items()},counts)
|
||||
for name,value in point.items():
|
||||
sample=boot[name];undefined=int(np.sum(~np.isfinite(sample)));lo,hi=(None,None) if undefined else np.percentile(sample,[2.5,97.5],method='linear').tolist();summary[name]=dict(value=float(value),low=lo,high=hi,undefined_replicates=undefined)
|
||||
result[metric].update(contrasts=summary,paired_component_ids=ids,paired_components=len(ids),expected_components=expected,complete_support=len(ids)==expected and expected>0,replicates=replicates,seed=seed,joint_draws_all_contrasts=True)
|
||||
if domain!='own_learning':
|
||||
field=result['relative_l2'];ci=field['contrasts'];result['field_classification']=classify(ci.get('s_2U',{}),ci.get('g_100',{}),complete=field['complete_support'])
|
||||
out[domain]=result
|
||||
return dict(domains=out,external_result_descriptive_reused_fold=True,uncertainty='Evaluation components only; no training seed/pool uncertainty',update_effect='Joint complete schedule and update budget change',no_training_selection=True)
|
||||
@@ -0,0 +1,88 @@
|
||||
"""Target-free endpoint prediction, then separately gated single-read scoring."""
|
||||
import json,os,sys,time
|
||||
from pathlib import Path
|
||||
import numpy as np
|
||||
from phase2_h1_worker_v2 import save,sha,checked
|
||||
from phase2_home_freeze_io import verify
|
||||
from phase2_gate_outcome_review import validate_review
|
||||
from phase3_item1_postfit_common import config,read,adopted_inputs,expected_checkpoint,manifest_identity,DOMAIN
|
||||
from phase3_item1_postfit_statistics import summarize
|
||||
from phase2_residual_math import decode_seconds
|
||||
from s2_custody_common import prediction_inputs
|
||||
from s2_custody_statistics import part_metrics
|
||||
|
||||
def query_indices(raw,q):
|
||||
ix=np.searchsorted(raw,q['raw_ids'])
|
||||
if np.any(ix>=len(raw)) or not np.array_equal(raw[ix],q['raw_ids']):raise ValueError('Fixed query/raw ID mapping')
|
||||
return ix
|
||||
|
||||
def save_prediction(root,arm,case,seconds,raw,q):
|
||||
ix=query_indices(raw,q);seconds=np.asarray(seconds,np.float64)
|
||||
if seconds.shape!=raw.shape or not np.isfinite(seconds).all():raise ValueError('Full-node prediction validity')
|
||||
p=root/'predictions'/arm/(case+'.npz');p.parent.mkdir(parents=True,exist_ok=True)
|
||||
if p.exists():raise FileExistsError('No prediction overwrite/retry')
|
||||
np.savez_compressed(p,seconds=seconds,raw_ids=raw,pred=seconds[ix],query_raw_ids=q['raw_ids'],counts=q['counts'],near_wall=q['near_wall'],coordinates_over_Rg=q['coordinates_over_Rg'])
|
||||
return dict(path=str(p),sha256=sha(p),nodes=len(raw),query_nodes=len(ix),negative_full_predictions=int(np.sum(seconds<0)),negative_query_predictions=int(np.sum(seconds[ix]<0)),origin='ITEM1_FROZEN_ENDPOINT_OR_CLOCK')
|
||||
|
||||
def predict(root,index):
|
||||
import torch
|
||||
from phase2_h2_model import H2Geo
|
||||
from phase2_h2_data import to_device,arguments,cached
|
||||
from phase3_item1_postfit_data import load_evaluation_part
|
||||
c,tc,pilot=config(root);validate_review(root);adopted=adopted_inputs(root,c,tc);m=adopted['models'][index];job=c['jobs'][index];f=tc['pools'][str(m['pool'])];view=dict(pilot,folds=[f]);out=root/'jobs'/str(index)/'prediction-records';out.mkdir(exist_ok=False);stats=read(m['standardization']);entries=[];parts=0;cpu=time.process_time();start=time.monotonic()
|
||||
if os.environ.get('CUDA_VISIBLE_DEVICES')!=c['gpu_uuid'] or os.environ.get('PYTORCH_CUDA_ALLOC_CONF')!='expandable_segments:True':raise ValueError('GPU allocator identity')
|
||||
if torch.__version__!=pilot['versions']['torch'] or np.__version__!=pilot['versions']['numpy']:raise ValueError('Library identity')
|
||||
os.sched_setaffinity(0,c['worker_cpus']);torch.set_num_threads(1);torch.set_num_interop_threads(1);torch.backends.cuda.matmul.allow_tf32=False;torch.backends.cudnn.allow_tf32=False;torch.cuda.set_device(0);torch.cuda.init();torch.cuda.reset_peak_memory_stats()
|
||||
def peaks():return dict(peak_allocated_bytes=int(torch.cuda.max_memory_allocated()),peak_reserved_bytes=int(torch.cuda.max_memory_reserved()))
|
||||
try:
|
||||
blob=torch.load(checked(m['checkpoint'],m['checkpoint_sha256']),map_location='cpu',weights_only=False);expected_checkpoint(job,blob,m['training_contract_sha256']);model=H2Geo().cuda();model.load_state_dict(blob['model'],strict=True);model.eval();del blob
|
||||
save(out/'environment.json',dict(torch=torch.__version__,gpu_uuid=c['gpu_uuid'],checkpoint_sha256=m['checkpoint_sha256'],allocator_settings=str(torch.cuda.memory._snapshot().get('allocator_settings',{})),dtype='FP32',autocast=False))
|
||||
old={e['case']:e for e in read(tc['S1_reuse']['test_manifest'])['entries']};emit_clock=index in [0,1,6]
|
||||
with torch.no_grad():
|
||||
for role,domain in DOMAIN.items():
|
||||
for case in f[role]:
|
||||
row=c['rows'][case];q=prediction_inputs(row);identity=dict(case=case,component_id=row['component_id'],pool=m['pool'],domain=domain)
|
||||
if index==6 and role=='test':
|
||||
e=old[case];checked(e['path'],e['sha256']);entries.append(dict(**identity,arm=m['arm'],**{k:e[k] for k in ['path','sha256','nodes','query_nodes','negative_full_predictions','negative_query_predictions']},origin='S1_HASH_REUSE'))
|
||||
else:
|
||||
data=to_device(load_evaluation_part(view,case,stats,role),'cuda');residual=model(*arguments(data),checkpoint_blocks=False)
|
||||
if not bool(torch.isfinite(residual).all()):raise ValueError('Nonfinite residual')
|
||||
seconds=decode_seconds(residual.cpu().numpy(),data['F'],f['c'],data['t_set']);torch.cuda.synchronize();entries.append(dict(**identity,arm=m['arm'],**save_prediction(root,m['arm'],case,seconds,data['raw_ids'],q)));del residual,data,seconds
|
||||
if emit_clock:
|
||||
raw=cached(view,case,'raw_ids');F=cached(view,case,'Vf');seconds=np.asarray(F,np.float64)*f['c']*pilot['parts'][case]['t_set'];arm='c'+str(m['pool'])+'F';entries.append(dict(**identity,arm=arm,**save_prediction(root,arm,case,seconds,raw,q)))
|
||||
parts+=1;save(out/'progress.json',dict(stage='PREDICTION_NO_TARGETS',job=index,parts=parts,entries=len(entries),wall_seconds=time.monotonic()-start,cpu_seconds=time.process_time()-cpu,**peaks()))
|
||||
save(out/'manifest.json',dict(entries=entries,targets_read=False));save(out/'result.json',dict(status='COMPLETE',parts=parts,entries=len(entries),targets_read=False,new_fits=0,wall_seconds=time.monotonic()-start,cpu_seconds=time.process_time()-cpu,**peaks()))
|
||||
except BaseException as exc:
|
||||
save(out/'failure.json',dict(status='FAILED',parts=parts,reason=type(exc).__name__+': '+str(exc),**peaks()));raise
|
||||
|
||||
def score(root):
|
||||
from phase2_h2_data import cached
|
||||
c,tc,pilot=config(root);validate_review(root);adopted_inputs(root,c,tc);verify(root,'predictions');entries=read(root/'predictions-manifest.json')['entries'];manifest_identity(entries,c,tc,pilot);out=root/'score';out.mkdir(exist_ok=False)
|
||||
save(out/'review-consumed.json',dict(unix=time.time(),review_sha256=sha(root/'review.json'),model_manifest_sha256=sha(root/'models-adopted.json'),prediction_manifest_sha256=sha(root/'predictions-manifest.json'),before_target_values=True))
|
||||
old=read(checked(tc['S1_reuse']['test_score'],tc['S1_reuse']['test_score_sha256']));oldrows={r['case']:r for r in old['rows']};f0=tc['pools']['100'];roles={case:role for role in DOMAIN for case in f0[role]};bycase={}
|
||||
for e in entries:bycase.setdefault(e['case'],[]).append(e)
|
||||
if set(bycase)!=set(roles) or len(bycase)!=552:raise ValueError('Complete common552 evaluation support')
|
||||
rows=[];access=[]
|
||||
for case in sorted(bycase):
|
||||
role=roles[case];q=prediction_inputs(c['rows'][case]);raw=cached(pilot,case,'raw_ids');ix=query_indices(raw,q);record=dict(case=case,role=role,target_sha256=pilot['parts'][case]['output_sha256']['seconds.npy'],unix=time.time())
|
||||
with (out/'target-access.jsonl').open('a') as log:log.write(json.dumps(record)+'\n')
|
||||
access.append(record);truth=cached(pilot,case,'seconds')
|
||||
for e in bycase[case]:
|
||||
with np.load(checked(e['path'],e['sha256']),allow_pickle=False) as z:
|
||||
if not np.array_equal(z['raw_ids'],raw) or not np.array_equal(z['query_raw_ids'],q['raw_ids']) or not np.array_equal(z['counts'],q['counts']) or not np.array_equal(z['near_wall'],q['near_wall']) or not np.array_equal(z['coordinates_over_Rg'],q['coordinates_over_Rg']) or not np.array_equal(z['pred'],z['seconds'][ix]):raise ValueError('Prediction/full/query identity')
|
||||
if len(z['seconds'])!=e['nodes'] or len(ix)!=e['query_nodes'] or int(np.sum(z['seconds']<0))!=e['negative_full_predictions'] or int(np.sum(z['pred']<0))!=e['negative_query_predictions']:raise ValueError('Negative/coverage metadata mismatch')
|
||||
if e['origin']=='S1_HASH_REUSE':
|
||||
oldrow=oldrows[case]
|
||||
if oldrow['component_id']!=e['component_id'] or oldrow['fold']!=0:raise ValueError('S1 stored metric identity')
|
||||
metrics=oldrow['new']
|
||||
else:metrics=part_metrics(z['pred'],truth[ix],q['counts'],q['near_wall'],q['raw_ids'],q['coordinates_over_Rg'],int(case.split('_')[0]))
|
||||
rows.append(dict(case=case,component_id=e['component_id'],arm=e['arm'],pool=e['pool'],domain=e['domain'],metrics=metrics,origin=e['origin']))
|
||||
del truth;save(out/'progress.json',dict(unique_target_reads=len(access),rows=len(rows)))
|
||||
if {r:sum(a['role']==r for a in access) for r in DOMAIN}!={'learning':320,'validation':111,'test':121}:raise ValueError('Exactly one read per frozen target/domain')
|
||||
comparison=read(c['comparison_contract']);result=summarize(rows,comparison['component_counts'],comparison['bootstrap']['replicates'],comparison['bootstrap']['seed'])
|
||||
negatives={arm:dict(full_nodes=sum(e['negative_full_predictions'] for e in entries if e['arm']==arm),query_nodes=sum(e['negative_query_predictions'] for e in entries if e['arm']==arm)) for arm in sorted({e['arm'] for e in entries})}
|
||||
save(out/'result.json',dict(status='COMPLETE',summary=result,rows=rows,unique_target_reads=len(access),test_target_reads=121,S1_metric_rows_reused=121,new_fits=0,new_decodes=0,negative_predictions=negatives,clipped=False,model_predictions_full_nodes=True,metrics='Original fixed PPS query readout; multiplicity once, query argmax last/Rg, original tau pairing',comparison_contract_sha256=c['comparison_contract_sha256'],review_sha256=sha(root/'review.json'),future_phase3_started=False))
|
||||
if __name__=='__main__':
|
||||
if sys.argv[1]=='predict':predict(Path(sys.argv[2]),int(sys.argv[3]))
|
||||
elif sys.argv[1]=='score' and len(sys.argv)==3:score(Path(sys.argv[2]))
|
||||
else:raise ValueError('predict root index or score root')
|
||||
@@ -0,0 +1,57 @@
|
||||
"""Prepare the one approved review during training; no target arrays or launches."""
|
||||
import json,shutil,time,re
|
||||
from pathlib import Path
|
||||
from phase2_h1_worker_v2 import sha,checked,save
|
||||
from phase2_h2_prepare import closure
|
||||
from phase2_home_freeze_io import verify,freeze
|
||||
from phase3_item1_postfit_common import read,config,MODEL_ARMS,expected_entries
|
||||
B=Path('/home/frank14f/optane/MeltMind')
|
||||
R=Path('/home/frank14f/MeltMind-runs/phase3-item1-postfit-2026-11-18')
|
||||
T=R.parent/'phase3-data-sufficiency-v2-2026-11-18'
|
||||
def main():
|
||||
import psutil
|
||||
start=time.process_time()
|
||||
if R.exists():raise FileExistsError('Existing postfit preparation: inspect, never duplicate')
|
||||
live=[]
|
||||
for p in psutil.process_iter(['pid','cmdline']):
|
||||
args=p.info['cmdline'] or []
|
||||
if any(Path(x).name in {'phase3_item1_postfit.py','phase3_item1_postfit_worker.py'} for x in args):raise RuntimeError('Existing postfit execution')
|
||||
if any(Path(x).name in {'phase3_data_sufficiency_v2_queue.py','phase3_data_sufficiency_v2_guard.py','phase3_data_sufficiency_v2_worker.py'} for x in args):live.append(dict(pid=p.pid,command=args))
|
||||
tp=verify(T,'implementation');tc=read(T/'contract.json')
|
||||
if tp['commit']!='73f94bbc77e5ed3156ec85ff4a32ec2c834fdb07' or sha(T/'contract.json')!='1bdf2331c5127e03bec160835eda9d04ff333c3f70d5fe12a0629ffa82ec5826':raise ValueError('Pinned training identity')
|
||||
sources=dict(tc['source_sha256'])
|
||||
for p,h in {**sources,**tc['code_sha256']}.items():checked(p,h)
|
||||
pilot=read(tc['pilot_contract']);hpath=B/'artifacts/phase2-h1-execution-2026-11-17/contract.json';rows=read(hpath)['grouped_rows'];s1train=B/'artifacts/phase2-h2-s1-2026-11-17/contract.json'
|
||||
s1=Path(tc['S1_reuse']['test_manifest']).parent;verify(s1,'predictions');verify(s1,'terminal')
|
||||
for p in [T/'contract.json',T/'implementation-git-freeze.json',T/'implementation-authorization.json',Path(tc['pilot_contract']),hpath,s1train,Path(tc['S1_reuse']['test_manifest'])]:sources[str(p)]=sha(p)
|
||||
old=read(tc['S1_reuse']['test_manifest'])['entries']
|
||||
if len(old)!=121 or {e['case'] for e in old}!=set(tc['pools']['100']['test']):raise ValueError('S1 test manifest coverage')
|
||||
for e in old:sources[e['path']]=e['sha256'];checked(e['path'],e['sha256'])
|
||||
jobs=[]
|
||||
for j,arm in zip(tc['jobs'],MODEL_ARMS):
|
||||
tag='epoch200' if j['mode']=='epoch' else ('U' if j['target_steps']==32000 else '2U')
|
||||
jobs.append(dict(id=j['id'],arm=arm,pool=j['pool'],mode=j['mode'],target_steps=j['target_steps'],tag=tag,expected_parts=sum(len(tc['pools'][str(j['pool'])][r]) for r in ['learning','validation','test'])))
|
||||
jobs.append(dict(id=6,arm=MODEL_ARMS[6],pool=100,mode='epoch',target_steps=64000,tag='last_epoch_only',expected_parts=552))
|
||||
universe=set().union(*(set(tc['pools']['100'][r]) for r in ['learning','validation','test']))
|
||||
if len(universe)!=552 or not universe<=set(rows):raise ValueError('Original evaluation roster')
|
||||
for pool in ['25','50','100']:
|
||||
f=tc['pools'][pool]
|
||||
if len({rows[x]['component_id'] for x in f['learning']})!=f['learning_components']:raise ValueError('Learning component counts')
|
||||
log=Path('/tmp/phase3-item1-postfit-tests3.log').read_text()
|
||||
if 'Ran 27 tests' not in log or not log.rstrip().endswith('OK'):raise ValueError('Actual27 tests PASS required')
|
||||
comparison=Path(tc['comparison_contract']);checked(comparison,'332f2515911c99aadf2486086ddff82bd195d889faa8391296443e7aeacdfee9')
|
||||
codes={str(B/p):h for p,h in closure(['phase3_item1_postfit','phase3_item1_postfit_common','phase3_item1_postfit_worker','phase3_item1_postfit_guard','phase3_item1_postfit_data','phase3_item1_postfit_statistics','prepare_phase3_item1_postfit','test_phase3_item1_postfit']).items()}
|
||||
doc=B/'docs/phase3-item1-postfit-2026-11-18.md';codes[str(doc)]=sha(doc)
|
||||
c=dict(version='phase3-item1-postfit-v1',approval='1794950613.195499',postfit_root=str(R),training_root=str(T),training_contract=str(T/'contract.json'),training_implementation=tp['commit'],S1_training_contract_sha256=sha(s1train),jobs=jobs,rows={x:rows[x] for x in sorted(universe)},source_sha256=sources,code_sha256=codes,comparison_contract=str(R/'comparison-contract.json'),comparison_contract_sha256=sha(comparison),postfit_gpu_cap=7200,postfit_cpu_cap=3600,item_gpu_cap=120000,gpu_uuid=tc['jobs'][0]['gpu_uuid'],worker_cpus=[0,1],coordinator_cpus=[12,13],new_fits=0,new_decodes=0,no_retry=True,no_auto_launch=True,review_stages=1,actual_checkpoint_hashes='Runtime only after six frozen terminals and process exit',all_predictions_frozen_before_target_read=True,expected_entries=4582,expected_unique_target_reads=552,expected_test_reads=121)
|
||||
expected=expected_entries(c,tc)
|
||||
if len(expected)!=4582:raise ValueError('Complete prediction cardinality')
|
||||
R.mkdir();shutil.copyfile(comparison,R/'comparison-contract.json');save(R/'contract.json',c)
|
||||
for attempt in [1,2,3]:shutil.copyfile('/tmp/phase3-item1-postfit-tests'+str(attempt)+'.log',R/('synthetic-attempt'+str(attempt)+'.log'))
|
||||
save(R/'acceptance.json',dict(status='PASS',tests=27,elapsed_seconds=float(re.search(r"Ran 27 tests in ([0-9.]+)s",log).group(1)),log_sha256=sha(R/'synthetic-attempt3.log'),test_source_sha256=sha(B/'tools/test_phase3_item1_postfit.py'),scope='CPU synthetic and isolated Git/file validator; no real targets, prediction, fits or CUDA workload',original18_pass_preserved=True))
|
||||
save(R/'training-live-snapshot.json',read(T/'live-progress.json'))
|
||||
save(R/'preparation-audit.json',dict(status='PREPARED_NOT_LAUNCHED',unix=time.time(),metadata_cpu_seconds=time.process_time()-start,training_implementation=tp['commit'],training_code_pins_verified=len(tc['code_sha256']),training_processes=live,training_complete_exists=(T/'complete.json').exists(),future_new_checkpoint_hashes_not_adopted=True,comparison_byte_identical=True,S1_prediction_hashes_checked=121,expected_prediction_entries=len(expected),model_entries=3265,new_model_entries=3144,reused_model_entries=121,clock_entries=1317,unique_parts=552,test_parts=121,new_fits=0,new_predictions=0,new_scores=0,test_target_values_read=False,old_score_values_read=False,training_untouched=True,new_review_pending=True,no_waiter_started=True))
|
||||
freeze(R,'implementation',[R/'contract.json',R/'comparison-contract.json',R/'acceptance.json',R/'preparation-audit.json',R/'training-live-snapshot.json',R/'synthetic-attempt1.log',R/'synthetic-attempt2.log',R/'synthetic-attempt3.log',*[Path(p) for p in codes]])
|
||||
proof=verify(R,'implementation');binding=dict(reviewed_commit=proof['commit'],contract_sha256=sha(R/'contract.json'),implementation_proof_sha256=sha(R/'implementation-git-freeze.json'),comparison_contract_sha256=c['comparison_contract_sha256'],code_sha256=codes);save(R/'expected-review-binding.json',binding)
|
||||
freeze(R,'review-binding',[R/'contract.json',R/'expected-review-binding.json',R/'implementation-git-freeze.json',R/'implementation-authorization.json']);config(R)
|
||||
print(json.dumps(dict(root=str(R),implementation=proof['commit'],contract_sha256=sha(R/'contract.json'),proof_sha256=sha(R/'implementation-git-freeze.json'),binding_sha256=sha(R/'expected-review-binding.json'),code_files=len(codes),review_pending=True,launched=False)))
|
||||
if __name__=='__main__':main()
|
||||
@@ -0,0 +1,86 @@
|
||||
import unittest,tempfile,ast,copy
|
||||
from pathlib import Path
|
||||
from unittest.mock import patch
|
||||
import numpy as np
|
||||
from phase3_item1_postfit_statistics import classify,contrasts,summarize,ARMS,METRICS
|
||||
from phase3_item1_postfit_common import expected_checkpoint,require_budget,terminal_inputs,terminal_identity,MODEL_ARMS,expected_entries
|
||||
from phase3_item1_postfit_worker import query_indices,score
|
||||
from test_phase2_gate_outcome_postfit import SemanticReview
|
||||
from phase3_item1_postfit_data import load_evaluation_part
|
||||
from phase3_item1_postfit_guard import final_reasons
|
||||
class PostfitTests(unittest.TestCase):
|
||||
def terminal_fixture(self):
|
||||
return (dict(status='TRAIN_COMPLETE_POSTFIT_PENDING',entries=[dict(id=i,status='PASS',returncode=0,gpu_seconds_charged=100) for i in range(6)],training_gpu_seconds=600,test_targets_read=False),dict(models=[dict(job=i) for i in range(6)],test_prediction_started=False,new_fits=6,new_checkpoints=6))
|
||||
def test_six_terminal_identity_and_once_ledger(self):
|
||||
d,m=self.terminal_fixture();self.assertEqual(terminal_identity(d,m),600)
|
||||
for change in ['partial','failed','duplicate','returncode','cost','nan','custody']:
|
||||
a,b=copy.deepcopy((d,m))
|
||||
if change=='partial':a['entries'].pop()
|
||||
if change=='failed':a['entries'][0]['status']='FAILED'
|
||||
if change=='duplicate':b['models'][0]['job']=1
|
||||
if change=='returncode':a['entries'][0]['returncode']=1
|
||||
if change=='cost':a['training_gpu_seconds']+=1
|
||||
if change=='nan':a['entries'][0]['gpu_seconds_charged']=float('nan')
|
||||
if change=='custody':a['test_targets_read']=True
|
||||
with self.assertRaises(ValueError):terminal_identity(a,b)
|
||||
def test_score_rejects_missing_PASS_before_targets(self):
|
||||
with tempfile.TemporaryDirectory() as d,patch('phase3_item1_postfit_worker.config',return_value=({}, {}, {})),patch('phase3_item1_postfit_worker.validate_review',side_effect=PermissionError('No actual new PASS')),patch('phase2_h2_data.cached') as cached:
|
||||
with self.assertRaises(PermissionError):score(Path(d))
|
||||
cached.assert_not_called();self.assertFalse((Path(d)/'score').exists())
|
||||
def test_score_rejects_missing_prediction_freeze_before_targets(self):
|
||||
with tempfile.TemporaryDirectory() as d,patch('phase3_item1_postfit_worker.config',return_value=({}, {}, {})),patch('phase3_item1_postfit_worker.validate_review'),patch('phase3_item1_postfit_worker.adopted_inputs'),patch('phase3_item1_postfit_worker.verify',side_effect=ValueError('Prediction hash mismatch')),patch('phase2_h2_data.cached') as cached:
|
||||
with self.assertRaises(ValueError):score(Path(d))
|
||||
cached.assert_not_called();self.assertFalse((Path(d)/'score').exists())
|
||||
def test_data_boundary(self):
|
||||
self.assertEqual(classify(dict(low=.001,high=.003),dict(low=0,high=.0005)),'UNIDENTIFIED_OR_OVERFIT_SIGNS')
|
||||
def test_four_classifications(self):
|
||||
pos=dict(low=.002,high=.004);eq=dict(low=-.001,high=.001)
|
||||
self.assertEqual(classify(pos,eq),'DATA_BENEFIT_AT_TESTED_RANGE');self.assertEqual(classify(eq,pos),'UPDATE_BUDGET_AND_SCHEDULE_BENEFIT');self.assertEqual(classify(pos,pos),'BOTH_BENEFITS');self.assertEqual(classify(eq,eq),'LOCAL_PLATEAU_ONLY')
|
||||
def test_wide_not_plateau(self):self.assertEqual(classify(dict(low=-.1,high=.1),dict(low=-.1,high=.1)),'UNIDENTIFIED_OR_OVERFIT_SIGNS')
|
||||
def test_incomplete_not_classified(self):self.assertEqual(classify(dict(low=.002,high=.003),dict(low=0,high=0),complete=False),'UNIDENTIFIED_OR_OVERFIT_SIGNS')
|
||||
def test_actual_count_denominator(self):
|
||||
a={k:.2 for k in ARMS};a['step100_2U']=.1;q=contrasts(a,{'25':34,'50':65,'100':132});self.assertAlmostEqual(q['s_2U'],.1/np.log2(132/65));self.assertNotEqual(q['s_2U'],.1)
|
||||
def rows(self):
|
||||
rows=[]
|
||||
for i in range(5):
|
||||
for arm in ARMS:
|
||||
value=(i+1)*.02+(0.003 if arm.startswith('step50') else 0)
|
||||
if arm in ['step50_2U','step100_2U']:value-=.002
|
||||
rows.append(dict(domain='common_test',arm=arm,case=str(i),component_id=str(i),metrics={k:dict(value=value,reason=None) for k in METRICS}))
|
||||
return rows
|
||||
def test_joint_draw_cancellation(self):
|
||||
s=summarize(self.rows(),{'25':34,'50':65,'100':132})['domains']['common_test']['relative_l2'];q=s['contrasts']['s_2U_minus_s_U'];self.assertAlmostEqual(q['low'],0);self.assertAlmostEqual(q['high'],0);self.assertTrue(s['joint_draws_all_contrasts'])
|
||||
def test_component_missing_not_silently_promoted(self):
|
||||
r=self.rows();r[0]['metrics']['relative_l2']=dict(value=None,reason='zero truth');s=summarize(r,{'25':34,'50':65,'100':132})['domains']['common_test'];self.assertEqual(s['relative_l2']['paired_components'],4);self.assertFalse(s['relative_l2']['complete_support']);self.assertEqual(s['field_classification'],'UNIDENTIFIED_OR_OVERFIT_SIGNS')
|
||||
def test_query_order_mapping(self):
|
||||
np.testing.assert_array_equal(query_indices(np.array([2,7,9]),dict(raw_ids=np.array([2,9]))),[0,2])
|
||||
with self.assertRaises(ValueError):query_indices(np.array([2,7,9]),dict(raw_ids=np.array([10])))
|
||||
def test_loader_denies_wrong_role_before_IO(self):
|
||||
with self.assertRaises(PermissionError):load_evaluation_part(dict(folds=[dict(learning=['a'],validation=['b'],test=['c'])]),'a',{},'test')
|
||||
def test_loader_no_target_branch(self):
|
||||
tree=ast.parse(Path('tools/phase3_item1_postfit_data.py').read_text());literals={n.value for n in ast.walk(tree) if isinstance(n,ast.Constant) and isinstance(n.value,str)};self.assertFalse({'target','seconds','target.npy','seconds.npy'}&literals)
|
||||
def test_prediction_no_label_calls(self):
|
||||
t=ast.parse(Path('tools/phase3_item1_postfit_worker.py').read_text());f=next(n for n in t.body if isinstance(n,ast.FunctionDef) and n.name=='predict');calls=[n for n in ast.walk(f) if isinstance(n,ast.Call) and isinstance(n.func,ast.Name) and n.func.id=='cached'];self.assertEqual({n.args[2].value for n in calls},{'raw_ids','Vf'})
|
||||
def test_checkpoint_identity(self):
|
||||
j=dict(id=2,target_steps=32000,pool=50,mode='step',tag='U');m=dict(fold=0,seed=0,optimizer_steps=32000,contract_sha256='x',pool=50,mode='step',role='U');expected_checkpoint(j,m,'x')
|
||||
for key,value in [('optimizer_steps',64000),('seed',1),('role','2U'),('contract_sha256','wrong')]:
|
||||
with self.assertRaises(ValueError):expected_checkpoint(j,dict(m,**{key:value}),'x')
|
||||
def test_S1_checkpoint_selection(self):
|
||||
j=dict(id=6,target_steps=64000);m=dict(fold=0,seed=0,optimizer_steps=64000,contract_sha256='old',completed_epochs=200,role='last_epoch_only');expected_checkpoint(j,m,'old')
|
||||
with self.assertRaises(ValueError):expected_checkpoint(j,dict(m,role='best_validation'),'old')
|
||||
def test_training_pending_refuses_before_adoption(self):
|
||||
with tempfile.TemporaryDirectory() as d:
|
||||
with self.assertRaisesRegex(RuntimeError,'WAIT_EXISTING'):terminal_inputs(dict(training_root=d),{})
|
||||
def test_budget_fixed_remaining(self):
|
||||
require_budget(112800,0,180,7)
|
||||
with self.assertRaises(RuntimeError):require_budget(112800,1000,180,7)
|
||||
with self.assertRaises(RuntimeError):require_budget(112800,0,600,7)
|
||||
def test_budget_nonfinite(self):
|
||||
with self.assertRaises(ValueError):require_budget(float('nan'),0,0,0)
|
||||
def test_failure_peaks_retained(self):
|
||||
class T:pass
|
||||
t=T();t.witnesses={'peak_allocated_bytes':{},'peak_reserved_bytes':{}};t.maxima={'peak_allocated_bytes':1,'peak_reserved_bytes':16*2**30}
|
||||
reasons=final_reasons([],dict(status='FAILED'),dict(status='FAILED',parts=1),t,20,20,300,1);self.assertIn('RECORDED_VRAM_LIMIT',reasons);self.assertIn('WORKER_NOT_PASS',reasons)
|
||||
def test_every_arm_expected_domain(self):
|
||||
c=dict(jobs=[dict(arm=a,pool=p) for a,p in zip(MODEL_ARMS,[25,50,50,50,100,100,100])]);tc=dict(pools={str(p):dict(learning=['l'+str(p)],validation=['v'],test=['t']) for p in [25,50,100]});e=expected_entries(c,tc);self.assertEqual(len(e),30);self.assertEqual(e[('S1_epoch100_reuse','t')]['domain'],'common_test')
|
||||
if __name__=='__main__':unittest.main()
|
||||
Reference in new issue
Block a user