diff --git a/.gitattributes b/.gitattributes index a6344aac8c09253b3b630fb776ae94478aa0275b..1dc8d47bac6518a358b49fce62e33d2c6df8e701 100644 --- a/.gitattributes +++ b/.gitattributes @@ -33,3 +33,5 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text *.zip filter=lfs diff=lfs merge=lfs -text *.zst filter=lfs diff=lfs merge=lfs -text *tfevents* filter=lfs diff=lfs merge=lfs -text +derived/reports/figures/evaluation_comparison.png filter=lfs diff=lfs merge=lfs -text +derived/reports/figures/training_curves.png filter=lfs diff=lfs merge=lfs -text diff --git a/HF_UPLOAD_MANIFEST.json b/HF_UPLOAD_MANIFEST.json new file mode 100644 index 0000000000000000000000000000000000000000..fcf0fa56601a53bd992c2e772d0a87d1b6ac25bb --- /dev/null +++ b/HF_UPLOAD_MANIFEST.json @@ -0,0 +1,799 @@ +{ + "artifact_path": "runs/mouse-run-run-2-mlp", + "created_at": "2026-07-03T19:17:33.391331+00:00", + "excluded_from_file_manifest": [ + "runs/mouse-run-run-2-mlp/HF_UPLOAD_MANIFEST.json" + ], + "file_count": 150, + "files": [ + { + "path": "README.md", + "sha256": "0a0de8947639614dfb361bee8756fd16ccbd89d188cd749394e973a910793024", + "size_bytes": 2898 + }, + { + "path": "cross_architecture_study/cross_arch_cka.json", + "sha256": "1df526c3c10435203135df1ba02285eef813ebbaa176995178168313b6aeb797", + "size_bytes": 518 + }, + { + "path": "cross_architecture_study/cross_arch_perunit.json", + "sha256": "fda1247aa82af44396e5ab786e57622b4699290b6daf61319860442b72a4a852", + "size_bytes": 14851 + }, + { + "path": "cross_architecture_study/cross_arch_plsc_residualized.json", + "sha256": "596160749242ca86dee143e914c6c981654fa85bedb024f355c0a3236a909f70", + "size_bytes": 4062 + }, + { + "path": "cross_architecture_study/figures/cross_arch_cka.png", + "sha256": "7b43e1f229f0cec33101c1b89634082a5b84445e45e85ee114be19ad0c9c3dbb", + "size_bytes": 60221 + }, + { + "path": "cross_architecture_study/figures/sharing_vs_vision.png", + "sha256": "b9ddfbb26313e9013310fe5f7236dc50a5bc7ba4af2e8dff9c90331a64ad4f2a", + "size_bytes": 74823 + }, + { + "path": "cross_architecture_study/report.md", + "sha256": "26dde4d9e0348308c638aafa454f34404b3b81667fdfa24b24f67d47f015bf71", + "size_bytes": 11540 + }, + { + "path": "derived/reports/MANIFEST.json", + "sha256": "d838078800d08a887708e013f6ce920ef429c889b58dbf5d2ee6c718f5e6ee45", + "size_bytes": 2710 + }, + { + "path": "derived/reports/figures/evaluation_comparison.png", + "sha256": "f884d82cdc4bac267de1ab48120986784960a6448d8c7f4502cc3b48c4369735", + "size_bytes": 172530 + }, + { + "path": "derived/reports/figures/training_curves.png", + "sha256": "f79e51dd791199e40a8057983a80cb9405bc95f186a2e5ee61eae94544f41bac", + "size_bytes": 386863 + }, + { + "path": "derived/reports/report.md", + "sha256": "df7d44d955304bc3e6c45960376cc9cbb01a1a6fd836ad63804117a491620e8c", + "size_bytes": 3022 + }, + { + "path": "derived/reports/summary.json", + "sha256": "585e4bebe1633bf2d7b81c3bf38b86097c9586976bcbb28591044cbdf32f5152", + "size_bytes": 14882 + }, + { + "path": "derived/tables/MANIFEST.json", + "sha256": "bfe73384c93185ec5f56aff90884b5b6088e608289dc7efa88755a86d38b8edd", + "size_bytes": 2327 + }, + { + "path": "derived/tables/evaluations.jsonl", + "sha256": "265ed5550a98149da5384ac6cbcd551db5b121176f7046a4ca58d89531a77cc7", + "size_bytes": 135821 + }, + { + "path": "derived/tables/exclusions.jsonl", + "sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "size_bytes": 0 + }, + { + "path": "derived/tables/panels.json", + "sha256": "cb73e84a077cbcd2d639c1c7f6947f3a418fa7482b62d1b789d5ad9608d30562", + "size_bytes": 760 + }, + { + "path": "derived/tables/rollouts.jsonl", + "sha256": "140147549967098f1a78dadab8e5888eb41a37389c094f79b623d22d7f0bd74a", + "size_bytes": 23011 + }, + { + "path": "derived/tables/runs.jsonl", + "sha256": "bfab762e5dd3673a8e08fa06635ebebd07bc02783f016fc26e56fee41756e2d8", + "size_bytes": 49200 + }, + { + "path": "eval.log", + "sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "size_bytes": 0 + }, + { + "path": "launch_gate.json", + "sha256": "6e2fd3b528fbd3efe3328e8f44e7679fa619599ea2ea8af098f034831db3ca65", + "size_bytes": 4762 + }, + { + "path": "logs/non_social_seed_0000_attempt_01.log", + "sha256": "3a2b6a56180126c0db00b005208128dc688d7176074287a417a6048706b6235b", + "size_bytes": 9880 + }, + { + "path": "logs/non_social_seed_0001_attempt_01.log", + "sha256": "0aa37b6b0cd019a3d6309d0b91244b5b9b5371e29a40dabecebb253465aff7a2", + "size_bytes": 9864 + }, + { + "path": "logs/non_social_seed_0002_attempt_01.log", + "sha256": "ccc2bd907f4f26a2f4cf78e398e77a5ba40edf05d24fbbee4c3b7dd8083765de", + "size_bytes": 9882 + }, + { + "path": "logs/non_social_seed_0003_attempt_01.log", + "sha256": "d476bca205d0fb8c0ef9b4c1d141feb1fbdf10bde7549ce9c4f4abe04707f7e9", + "size_bytes": 9881 + }, + { + "path": "logs/non_social_seed_0004_attempt_01.log", + "sha256": "907770647fa99e4811ead15c437c5e168abda4ec9add43af615dfbb0d0f58b15", + "size_bytes": 9891 + }, + { + "path": "logs/non_social_seed_0005_attempt_01.log", + "sha256": "bf9cd3e1fd692ccf839ff21ef79ce5df0d509055b20bd875fd1b316ce53f0472", + "size_bytes": 9890 + }, + { + "path": "logs/non_social_seed_0006_attempt_01.log", + "sha256": "c246f36daaadde9b5b4d7b15dbd746d8709be8f3eecc5e571d457e004e0684bd", + "size_bytes": 9884 + }, + { + "path": "logs/non_social_seed_0007_attempt_01.log", + "sha256": "cd66e9a39b49a3fa81a06659bd3deaf83f1633d4e703bc8fee128ed917843e77", + "size_bytes": 9891 + }, + { + "path": "logs/non_social_seed_0008_attempt_01.log", + "sha256": "5dff30a9bf230401173ee2d01636d2aa93de7acabf658ad63293bfe7e6b41452", + "size_bytes": 9890 + }, + { + "path": "logs/non_social_seed_0009_attempt_01.log", + "sha256": "dea4f6c6a2089721690d705414c31c73109f620b10c6389f2f97752fd20ed2be", + "size_bytes": 9897 + }, + { + "path": "logs/social_seed_0000_attempt_01.log", + "sha256": "2986b8e85c9e56c87c016f21bc310c04d7d8116c19ba77f77db5bc3be2bb6e15", + "size_bytes": 9898 + }, + { + "path": "logs/social_seed_0001_attempt_01.log", + "sha256": "3455734d414ab8fbdc514ef9048e02b5bd395c48c06caa02fabce4c29fc68aa4", + "size_bytes": 9885 + }, + { + "path": "logs/social_seed_0002_attempt_01.log", + "sha256": "1554d4b94d8f373fbc07624ec01a5da600d7fe5ba0e20e3f79366658da1c8ac1", + "size_bytes": 9885 + }, + { + "path": "logs/social_seed_0003_attempt_01.log", + "sha256": "19c661ccbcfa51b889b754fe2889d7b16900a1c74e5a7bd921701005eb5e6c31", + "size_bytes": 9883 + }, + { + "path": "logs/social_seed_0004_attempt_01.log", + "sha256": "32b92be23199e2511a7bb487aed2ba1ae774858f4bdd3a00b262191e1da1706e", + "size_bytes": 9880 + }, + { + "path": "logs/social_seed_0005_attempt_01.log", + "sha256": "693f324e857090798faa9e31bfaa8f68f0ff52315a5019ba85dd225886a8e1ae", + "size_bytes": 9895 + }, + { + "path": "logs/social_seed_0006_attempt_01.log", + "sha256": "2a96ed1152e16dbf9cd9be397c102db33665cc3c8289d3ef3735e72af8e39186", + "size_bytes": 9900 + }, + { + "path": "logs/social_seed_0007_attempt_01.log", + "sha256": "846fdea329feb45345bd55b3059f4eaf5dd50c3e8a94296e14de8ce2f9efb6f0", + "size_bytes": 9898 + }, + { + "path": "logs/social_seed_0008_attempt_01.log", + "sha256": "92b4b1dacc715edec48941233f5b96dd2a3333b1850375122e4331c68285e170", + "size_bytes": 9872 + }, + { + "path": "logs/social_seed_0009_attempt_01.log", + "sha256": "a3ed2eae88f29997826bc60c90a32a6762cbd5cb32c214f5049a36257291d00f", + "size_bytes": 9860 + }, + { + "path": "manifest.json", + "sha256": "5bd2d31615b52c6a71b1e429aaa8700ed125ba9f9f1a530b8f33b8fe49925f8e", + "size_bytes": 14522 + }, + { + "path": "non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "sha256": "f8e5409f712499d04d8944b4ecd7070e6a1688c886ed4463cf04bf85fc9707a3", + "size_bytes": 11460376 + }, + { + "path": "non_social/seed_0000/attempt_01/config.json", + "sha256": "d0dd81d4e5d43bbad6fb2fb66c9fb53c3c58778d489ae11c0d278e3ac12abb98", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0000/attempt_01/metrics.jsonl", + "sha256": "f22bc9638550a95a739113e80fb0fab418aec865b6828d2c31e958688155a1b5", + "size_bytes": 54203 + }, + { + "path": "non_social/seed_0000/attempt_01/status.json", + "sha256": "671947ec08a3ba606c15d40fdfc2d648def59e8e0ade1ba19bd1647875276390", + "size_bytes": 1367 + }, + { + "path": "non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "sha256": "a6c8227e6a0fa824f91e1ee406af15f03ca2f65ee00d67445626986d7698c1c3", + "size_bytes": 11460368 + }, + { + "path": "non_social/seed_0001/attempt_01/config.json", + "sha256": "d80a42266c7a8dc06278f223dc4856658290ca6ca5de3b4d2db5e43084d98185", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0001/attempt_01/metrics.jsonl", + "sha256": "b3835354e5a75760d24115eef1ed938698d8f8c183687fea31af81c43b9a44d0", + "size_bytes": 54155 + }, + { + "path": "non_social/seed_0001/attempt_01/status.json", + "sha256": "f6dd4b15cd3e20fc0cd83bc84cec04dd038d36480b330f78a55caabb5f27caf1", + "size_bytes": 1358 + }, + { + "path": "non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "sha256": "e94060644c4de0c5467f9d15df5a5e4cfc6fbff18dcf358eb5cebe1dce94d3d7", + "size_bytes": 11460384 + }, + { + "path": "non_social/seed_0002/attempt_01/config.json", + "sha256": "f7bb8be082e2f25f89d008e4d0e3b9630dd24dce7af045a4018e95750b5f7a3e", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0002/attempt_01/metrics.jsonl", + "sha256": "a9283e7b6dbe7047f9ba222b9051cdeb19dbaa86198a68ca512710b21f94d7c5", + "size_bytes": 54093 + }, + { + "path": "non_social/seed_0002/attempt_01/status.json", + "sha256": "293c9cda446c7ea2f8c1cedff7793f052b70fe3abc31bd8c8758b026efb77539", + "size_bytes": 1371 + }, + { + "path": "non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "sha256": "17f68c1c23df5485beda4cdb0392ee32910c36897733023f84588c380da48191", + "size_bytes": 11460384 + }, + { + "path": "non_social/seed_0003/attempt_01/config.json", + "sha256": "0380ad17a8054b1c662e8cac4df015a96649e416dd5dc29eb0f46aff45917d5d", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0003/attempt_01/metrics.jsonl", + "sha256": "9ca6e1775ba7ef2367f12aadafe906e78c6a47a168f81bb08cb5f3f20678a1e9", + "size_bytes": 54134 + }, + { + "path": "non_social/seed_0003/attempt_01/status.json", + "sha256": "794957f5ae2c8c7b84da03deb9b15cd7b2492bff9fc701d850ffcce80c026b5a", + "size_bytes": 1374 + }, + { + "path": "non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "sha256": "1fa11ee31375fa76d13192833c9376692c520fcc925fee9c6b76a18abd4318fa", + "size_bytes": 11460392 + }, + { + "path": "non_social/seed_0004/attempt_01/config.json", + "sha256": "23629c5b243e4ce830a736a5e027e0f3d091a167e6c71bb0a58d4660dce00005", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0004/attempt_01/metrics.jsonl", + "sha256": "2ce1f9f59a65048a7b6f14af65e29d9ee9ac017471cf73a73f4bd24a159ca207", + "size_bytes": 54115 + }, + { + "path": "non_social/seed_0004/attempt_01/status.json", + "sha256": "d06cf0ff249fbe63f7a9a3cd31b5478ff83dca66ed511cf91310bc985aa4416c", + "size_bytes": 1380 + }, + { + "path": "non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "sha256": "5debd7cc50f687b861a1987d4433d6434b53e3ca6cd4c4743ac9192794e9ed0c", + "size_bytes": 11460392 + }, + { + "path": "non_social/seed_0005/attempt_01/config.json", + "sha256": "d14d2d282339aef9695bfdcfcd9323ba34a14012cba8d22f4c9bf8d3470c8986", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0005/attempt_01/metrics.jsonl", + "sha256": "23bb5e4f45b3bd07dde09c4e377826d37627fe941bcbd6a51d1c1a2086363efa", + "size_bytes": 54230 + }, + { + "path": "non_social/seed_0005/attempt_01/status.json", + "sha256": "40bd156060de2d8959b1160ed642e1430d337e6ab9243f748a2e0f5cb5766de8", + "size_bytes": 1383 + }, + { + "path": "non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "sha256": "01b8815629b4e75ed7baddd26bfade4a209f1c7c07e34b4bc26af47afcf2b7ed", + "size_bytes": 11460384 + }, + { + "path": "non_social/seed_0006/attempt_01/config.json", + "sha256": "65872e4b0c3720b0f541200245f545fc070a4e5c59ccad3488a4d2dc97510da1", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0006/attempt_01/metrics.jsonl", + "sha256": "b646d4a29dcb91f9901b4e4fa58b8b5b6e85e488d790251d69831f8449e3bdcb", + "size_bytes": 54179 + }, + { + "path": "non_social/seed_0006/attempt_01/status.json", + "sha256": "e483ebdb019fe6734731b01abf488a79ae8d49fb495e1695a5ba9a67fa7a511c", + "size_bytes": 1375 + }, + { + "path": "non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "sha256": "556bd8154b82d372e4f881e6319eeee0a1ad15f4e6c1b04df358d0711c48283b", + "size_bytes": 11460392 + }, + { + "path": "non_social/seed_0007/attempt_01/config.json", + "sha256": "41c29face687bbafdc78400210bb26a99b7c786300c9422ebcb4ea15eff03d32", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0007/attempt_01/metrics.jsonl", + "sha256": "2914bb836de2d92e9802974bc62aa62e848dd5d662de22743fde454e853270ed", + "size_bytes": 54109 + }, + { + "path": "non_social/seed_0007/attempt_01/status.json", + "sha256": "370e92fc6865eef1c1633b063798462e5a730f88b14ed595cd9ebdd193b1d894", + "size_bytes": 1384 + }, + { + "path": "non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "sha256": "d5ff79b8953d863166e3771c86afe125054881d88328fa27cd20f2770011ed51", + "size_bytes": 11460392 + }, + { + "path": "non_social/seed_0008/attempt_01/config.json", + "sha256": "9f4fd591b08685dbfb5dd5f6d91c4cdc9ba96a2b46e0163806a9135a7af8da0d", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0008/attempt_01/metrics.jsonl", + "sha256": "d80c0a1058016565bd18b54280b6cb7e5e87d399887079bcfa68813e1befca0d", + "size_bytes": 54226 + }, + { + "path": "non_social/seed_0008/attempt_01/status.json", + "sha256": "e79c1a7b7dc8db65c6d5b1b46713ccf52697d089b7a1f110355847c7e0f17c64", + "size_bytes": 1378 + }, + { + "path": "non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "sha256": "def8dbb315937927f72b78afdd9894dd5613ef88bc1bdbdb7b654ba5e17e690d", + "size_bytes": 11460400 + }, + { + "path": "non_social/seed_0009/attempt_01/config.json", + "sha256": "709f0eccca00218f98572be643caa8499ce9c3f2ba86cfeb3eedf3496f595209", + "size_bytes": 1185 + }, + { + "path": "non_social/seed_0009/attempt_01/metrics.jsonl", + "sha256": "b6d2c2e74baa3116060ab1bd83dc02fea726165ad25f4d91c1524a063211b281", + "size_bytes": 54154 + }, + { + "path": "non_social/seed_0009/attempt_01/status.json", + "sha256": "5932b45256cb8254b89d1be4fdd76e50a17d3b96cf11ccb2922223ae5a8d7559", + "size_bytes": 1386 + }, + { + "path": "paper_random_eval.jsonl", + "sha256": "287c7b3cbf320a6e79a47ff80c4bc086a7077076abe6ef0b800ebc663d2e6a3b", + "size_bytes": 132390 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors", + "sha256": "47235ca3adb8a5800cf9afc150039b2a465987c50ff85daa8b9f7e2eb91b4700", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors", + "sha256": "e6dcd1a1b640313bf009d7c3af192d2acfd254fa7e355e524c6f8ba35c4eae0a", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors", + "sha256": "6004b9a96caddd23e6c5370cb3a7dcbc79e7ccf86415bfc293f4f2859c725039", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors", + "sha256": "a26129d1634332982cfc7436c50c3b54df91edd77569ff235c1a0d8f18507383", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors", + "sha256": "8001169fe126091e26b133edb59a13bb3bff90ab5826b6e32dd7745102385788", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors", + "sha256": "620dbd11a3b2f5485389a71091b2f45458b317db672910f107c567a98f03c28f", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors", + "sha256": "cb1d4e549c50e6326007a139e6491233b34389321ae2fbab23d58825b7dbd8d4", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors", + "sha256": "a941f04ea8dcce75abf0a42ed2329f1c987af0a7158fd5f92592eb67d418386e", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors", + "sha256": "3c249dac8de670dffeaf04b791c3b76f1f7015fad519831221eaf905d9e1b668", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors", + "sha256": "ecd6db0f030df8ca8995578bb1416a8d00c4293749b545e0dbf7e583909903d4", + "size_bytes": 46506651 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors", + "sha256": "0bad577c731d05df8b417b04bccfce215674bf0397bf0c87f6ea3f629e373f67", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors", + "sha256": "0e681e5199bb74bbb486c17420e6531bde8ca7f4873ba857ffd839b9e2e0bd1f", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors", + "sha256": "2c1daa497f3a0b5503f0e5f7a87fea538c6dc3407fc2adf0dfa6d26bc726d5b4", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors", + "sha256": "ceb7f1689016b609d5e2713a0a2661ca9f83078a5aa89dd66851f17525c1f890", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors", + "sha256": "c12193945e93692204781680d7af3c920b1a0307a09e31072266eb2a5e4610db", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors", + "sha256": "01465d07d01ed57f5c00a757c8a01ed8d23ccab455e35db5ae547eff48410011", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors", + "sha256": "6433519f01ecece7e0023cf22dd156df34773a4064fd7e051275a0b34b41510f", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors", + "sha256": "86ef95b85618bcb31066bdd41f692e9b045da238f8e252197743daff8da6ac17", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors", + "sha256": "76c5e4773e0a78d321490080d05ddc09f3af942e00d6c8958cdb4e3d7cefcea4", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors", + "sha256": "e98fced65d944fcdfc904a1b8344be4caaa933f30cb055861f623183b3df5dc9", + "size_bytes": 46506635 + }, + { + "path": "paper_rollouts/paper_rollout_records.jsonl", + "sha256": "bbe20472146213ba809436a6b0f28bfdcdad62b9e0f82bccd15d5e91346584c1", + "size_bytes": 19880 + }, + { + "path": "raw_records.jsonl", + "sha256": "be1e1692b0034a17180ddfc25d05bb02189b1cb558bd37060f90b8e283780445", + "size_bytes": 44596 + }, + { + "path": "report.log", + "sha256": "d838078800d08a887708e013f6ce920ef429c889b58dbf5d2ee6c718f5e6ee45", + "size_bytes": 2710 + }, + { + "path": "rollout.log", + "sha256": "0f815edad73fd2a4798725c22e7e2a8e2742646514fbc33277e9e46052dbac3f", + "size_bytes": 3060 + }, + { + "path": "run_status.json", + "sha256": "ef25ee077c79057a1fa041cbdb82bb7e7613546f5660088a82fb81ed3100f322", + "size_bytes": 38837 + }, + { + "path": "runner.log", + "sha256": "fbe8db5bee1737a5974e26e0fa79cdb31a4758bd9d61271010435804e07d6436", + "size_bytes": 3160 + }, + { + "path": "social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "sha256": "f4bcf1374937905b5d1f57c73129408305c6289a4aa0a054328102a1a4ceded0", + "size_bytes": 11460408 + }, + { + "path": "social/seed_0000/attempt_01/config.json", + "sha256": "26cc60a74e1f3e986a586451e1fbde856745fd6e48fba88062d1db3549d5fbd7", + "size_bytes": 1172 + }, + { + "path": "social/seed_0000/attempt_01/metrics.jsonl", + "sha256": "ced0dd7834936ea5020eda15921b74b58aafd9bfafde8e230083d97bc1055f23", + "size_bytes": 55340 + }, + { + "path": "social/seed_0000/attempt_01/status.json", + "sha256": "2e67a9fbca51f419e401e3cfff3ff4cc328bef207978a577e125df5f5e8f8c59", + "size_bytes": 1409 + }, + { + "path": "social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "sha256": "099cb185f62dd1f3f5f4d2842a1b0d283ade548522932f999c600e73c1ff44a8", + "size_bytes": 11460400 + }, + { + "path": "social/seed_0001/attempt_01/config.json", + "sha256": "184f4fc02e62aa29bc0d032edc7a322ae5c43832eab6a02947192b57c683b27a", + "size_bytes": 1172 + }, + { + "path": "social/seed_0001/attempt_01/metrics.jsonl", + "sha256": "ac0c73f137ab5242bcbb03dfdb981bb518828bfbaf2c1dffa44105d8db91021a", + "size_bytes": 55115 + }, + { + "path": "social/seed_0001/attempt_01/status.json", + "sha256": "cba00b2946080137b5be3675400144d40cdccbb1fab53fb153f5fd52770f1fb2", + "size_bytes": 1397 + }, + { + "path": "social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "sha256": "ddabd74e45c2c760a304a502299b61ff7db69c634e36f1f146feb9bf84e8fa5c", + "size_bytes": 11460400 + }, + { + "path": "social/seed_0002/attempt_01/config.json", + "sha256": "0b72006270f31460cf0317a5cc158b490fced4e44d7684a1ef9fdb560e4d749d", + "size_bytes": 1172 + }, + { + "path": "social/seed_0002/attempt_01/metrics.jsonl", + "sha256": "113a751b12158879e5b2b4f7500575c7f91f74edca88c16fcb2c5ab3cdcd7deb", + "size_bytes": 55101 + }, + { + "path": "social/seed_0002/attempt_01/status.json", + "sha256": "c1811a72c09743636b3bf2c3d657a00e0b827a5086f9dbe11d0b62abc82e1c92", + "size_bytes": 1398 + }, + { + "path": "social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "sha256": "77872aaef8ab25b68bf6d92a4af4bf74e5cc79b997d040bc34175db56af8e630", + "size_bytes": 11460400 + }, + { + "path": "social/seed_0003/attempt_01/config.json", + "sha256": "594facabfc3dad9237cbdb1bd1b0a4a87d7cb0cd6bd04d46ddf5f9179687f6de", + "size_bytes": 1172 + }, + { + "path": "social/seed_0003/attempt_01/metrics.jsonl", + "sha256": "048013e9095fe71db0cdf825b2d3d7acbb71e45af6a839527149e58023929d0e", + "size_bytes": 55169 + }, + { + "path": "social/seed_0003/attempt_01/status.json", + "sha256": "de6ff3ad58a9bbad450d69a2fef2a96ccb7f37277e0cb11356e313543ecd63ef", + "size_bytes": 1400 + }, + { + "path": "social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "sha256": "5f51e56c6790b8377549c656b1f50cce6f1bc65fa30c4ac612097554d1d0555f", + "size_bytes": 11460400 + }, + { + "path": "social/seed_0004/attempt_01/config.json", + "sha256": "66262f8aedc27441114a24a640b40c24428646f9b02e553a52e734df7210697f", + "size_bytes": 1172 + }, + { + "path": "social/seed_0004/attempt_01/metrics.jsonl", + "sha256": "5eb3af04e2ab4ab6bb5ac2c919c8a0cead0094136fd80aa8121aae6dbcda2b52", + "size_bytes": 55228 + }, + { + "path": "social/seed_0004/attempt_01/status.json", + "sha256": "7815b8148c28caa8e9b5bb63c088d64e7c2a87db14c9dab044f4d4a500926056", + "size_bytes": 1396 + }, + { + "path": "social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "sha256": "3405d3fc0a24dd17d23126ea788a9e1e3877444eb16623f74e0636865ace9e9b", + "size_bytes": 11460408 + }, + { + "path": "social/seed_0005/attempt_01/config.json", + "sha256": "416015f50c69ffef39d25e76049644951b1165dc8348c3b0353b8c78df1e65d2", + "size_bytes": 1172 + }, + { + "path": "social/seed_0005/attempt_01/metrics.jsonl", + "sha256": "20148b7f959cb2a43bc3e9eb513af5fe0339b16d32301243806b96092e6515b5", + "size_bytes": 55264 + }, + { + "path": "social/seed_0005/attempt_01/status.json", + "sha256": "23cf4d86baf4c7c4363e90431bfb5fd9f8733b1660a7154b27013a42db862f71", + "size_bytes": 1409 + }, + { + "path": "social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "sha256": "6a45c928ba31f02d9b235a4605d844df712be9103fbf93bfe039029aa9400fc4", + "size_bytes": 11460416 + }, + { + "path": "social/seed_0006/attempt_01/config.json", + "sha256": "7c5d18581b89ca5df4c1d4be8ebe8e9ba37fccc43ea3726c3dc8c97b795ceaab", + "size_bytes": 1172 + }, + { + "path": "social/seed_0006/attempt_01/metrics.jsonl", + "sha256": "516cf1adbc9c0075b8d9ebf320cd055f3d4b1d87db5376b1688db84120d57840", + "size_bytes": 55234 + }, + { + "path": "social/seed_0006/attempt_01/status.json", + "sha256": "97c600750a655b3c59a2eb1148d26f38e4ae8dfce850f91a71bf74b0726348fe", + "size_bytes": 1413 + }, + { + "path": "social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "sha256": "cb8a8e1bcbbb446b3ce500375218bbb11f70265bebe4d228299de67891943d15", + "size_bytes": 11460408 + }, + { + "path": "social/seed_0007/attempt_01/config.json", + "sha256": "1c3afe0ad677f8ae4c0ba672e2a4e76cd4b6c4f8857c0784cafb10b08a911944", + "size_bytes": 1172 + }, + { + "path": "social/seed_0007/attempt_01/metrics.jsonl", + "sha256": "4b67cc780716c9ef5557548c3396fed808c8b874ba5251b49412ee395ca8d39d", + "size_bytes": 55253 + }, + { + "path": "social/seed_0007/attempt_01/status.json", + "sha256": "f1676b961b553176b60650a6f26c591343b93c0f2b903d8eb311aea11dded755", + "size_bytes": 1408 + }, + { + "path": "social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "sha256": "39e2f67475bf8a4eca5f24de315185f9f759d1f65ac80f68a5644a58f16272e1", + "size_bytes": 11460392 + }, + { + "path": "social/seed_0008/attempt_01/config.json", + "sha256": "556e53904cd5c3e266faec185887090adf5d21612f4be7ad318482c0aafb871d", + "size_bytes": 1172 + }, + { + "path": "social/seed_0008/attempt_01/metrics.jsonl", + "sha256": "4e66fffd1375d22dbb32b2ec23e46b4b9519a9bf1cbcdf27e9104b740eaac813", + "size_bytes": 55240 + }, + { + "path": "social/seed_0008/attempt_01/status.json", + "sha256": "850d49a6859a22f969c245bdbea4c3a155a8b02fc6a3a7a4e90e49316130e5e9", + "size_bytes": 1387 + }, + { + "path": "social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "sha256": "6790e80fe441a2995b3c9817361ed5add9b0fbc14ca34187b08685868aad6875", + "size_bytes": 11460376 + }, + { + "path": "social/seed_0009/attempt_01/config.json", + "sha256": "71afc05bd0004920ad0b97efb94bc3792c6ace32a2f13ccd3e1f8a39cba16d7d", + "size_bytes": 1172 + }, + { + "path": "social/seed_0009/attempt_01/metrics.jsonl", + "sha256": "08b517e6e73bc0d339c0166dbca4d8ee03e1ffd4b176c5bd30278ee2b9bb930e", + "size_bytes": 55147 + }, + { + "path": "social/seed_0009/attempt_01/status.json", + "sha256": "8039fd890ceb6927a99458c8ec08f1f466eef28e2dd18ea2f30d861da9043807", + "size_bytes": 1373 + }, + { + "path": "tables.log", + "sha256": "bfe73384c93185ec5f56aff90884b5b6088e608289dc7efa88755a86d38b8edd", + "size_bytes": 2327 + }, + { + "path": "triton_equivalence.json", + "sha256": "6fa7c8f5a90df05c0af5f7759f369fea3a0e468f998353df8f6f8b1b91a2ad99", + "size_bytes": 1935 + } + ], + "path_in_repo": "", + "private": false, + "provenance": { + "command": "/Users/jacoblincool/Documents/GitHub/mouse-run-run/.venv/bin/upload-marl-artifacts runs/mouse-run-run-2-mlp --repo-id JacobLinCool/mouse-run-run-2-mlp --repo-type model --commit-message 'Upload mouse-run-run-2-mlp: trained agents, rollouts, evaluation, cross-architecture study'", + "created_at": "2026-07-03T19:17:33.857630+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": null + }, + "executable": "/Users/jacoblincool/Documents/GitHub/mouse-run-run/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "043425050d6ebb6201f0e8be035085405cb044ea", + "short_sha": "0434250", + "status_short": "" + }, + "platform": "macOS-26.4-arm64-arm-64bit-Mach-O", + "python": "3.14.4 (main, Apr 14 2026, 14:46:33) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": false, + "cuda_device_count": 0, + "cuda_devices": [], + "cuda_version": null, + "cudnn_version": null, + "mps_available": true, + "version": "2.12.1" + } + }, + "repo_id": "JacobLinCool/mouse-run-run-2-mlp", + "repo_type": "model", + "revision": null, + "schema_version": 1 +} diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..d5df8a695028627a5f2dc8a56d31efe9e0728801 --- /dev/null +++ b/README.md @@ -0,0 +1,71 @@ +--- +license: mit +tags: +- multi-agent-reinforcement-learning +- ppo +- recurrent-neural-network +- neuroscience +- inter-brain-dynamics +--- + +# mouse-run-run — MLP agents (mouse-run-run-2-mlp) + +Trained chaser/explorer agents for a modern PyTorch reproduction and +architecture extension of the MARL experiment in Zhang et al. (2025), +*Inter-brain neural dynamics in biological and artificial intelligence +systems*, Nature 645, 991–1001. + +**Architecture:** frame-stack MLP (k=8, memoryless — common-input floor control). Analysis (hidden) dimension 256; two independently +parameterized actor-critic agents (chaser and explorer, no weight sharing). + +This repo holds one full experiment: **20 trained agent pairs** (tasks +`social` and `non_social` × seeds 0–9), each trained with PPO for 20,000 +updates × 40 episodes × 100 steps. 20/20 units completed on the first attempt. + +## Contents + +- `{social,non_social}/seed_XXXX/attempt_01/checkpoints/latest.safetensors` — + trained chaser+explorer weights (safetensors; config/metrics in metadata). +- `paper_rollouts/*.safetensors` — 25×500-step analysis rollouts per pair: + per-timestep hidden states, positions, actions, and social-event flags + (schema v3, time-aligned; degenerate episodes flagged). +- `paper_random_eval.jsonl` — standardized random-opponent evaluation. +- `derived/{tables,reports}/` — canonical tables and the per-experiment report. +- `cross_architecture_study/` — the four-architecture comparison report, + figures, and cross-architecture PLSC/CKA data. +- `manifest.json`, `raw_records.jsonl`, `launch_gate.json`, + `triton_equivalence.json` — provenance. + +## Key results (this architecture) + +Behavior (vs standardized random opponent, social task): +- chaser collisions / episode: **26.93** + (non-social control: 2.68) +- chaser partner-in-vision: 0.93 + +Neural (social agents): +- collision decoding (balanced acc): **0.91** +- partner escape / approach decoding: 0.78 / 0.78 +- PLSC shared-dimension top correlation: 0.68 + +See `cross_architecture_study/report.md` for the four-architecture comparison. +**Headline:** the architectures do not learn the same internal representations — +only the RNN develops genuine internal shared dynamics at low mutual vision. + +## Load a checkpoint + +```python +from huggingface_hub import hf_hub_download +from mouse_run_run.serialization import load_checkpoint +from mouse_run_run.policy import build_policy + +path = hf_hub_download("JacobLinCool/mouse-run-run-2-mlp", + "social/seed_0000/attempt_01/checkpoints/latest.safetensors") +config, metrics, chaser_state, explorer_state = load_checkpoint(path) +chaser = build_policy(config["architecture"], config["env"]["observation_size"], + hidden_size=config["hidden_size"]) +chaser.load_state_dict(chaser_state) +``` + +Code: https://github.com/JacobLinCool/mouse-run-run (paper reproduction + +cross-architecture study). diff --git a/cross_architecture_study/cross_arch_cka.json b/cross_architecture_study/cross_arch_cka.json new file mode 100644 index 0000000000000000000000000000000000000000..831a30d3f93b9fafe16917f0da5c5d18c63cc2fc --- /dev/null +++ b/cross_architecture_study/cross_arch_cka.json @@ -0,0 +1,18 @@ +{ + "RNN|RNN": 1.0, + "RNN|MLP": 0.6792286610675173, + "RNN|SSM": 0.6359547814617931, + "RNN|Transformer": 0.6660120097138259, + "MLP|RNN": 0.6792286610675173, + "MLP|MLP": 1.0, + "MLP|SSM": 0.7110754305522333, + "MLP|Transformer": 0.7809858975192786, + "SSM|RNN": 0.6359547814617931, + "SSM|MLP": 0.7110754305522333, + "SSM|SSM": 1.0, + "SSM|Transformer": 0.7551091733153037, + "Transformer|RNN": 0.6660120097138259, + "Transformer|MLP": 0.7809858975192786, + "Transformer|SSM": 0.7551091733153037, + "Transformer|Transformer": 1.0 +} \ No newline at end of file diff --git a/cross_architecture_study/cross_arch_perunit.json b/cross_architecture_study/cross_arch_perunit.json new file mode 100644 index 0000000000000000000000000000000000000000..704f0c3008ce30abcfccfdca3a8f899ab84ecc24 --- /dev/null +++ b/cross_architecture_study/cross_arch_perunit.json @@ -0,0 +1,802 @@ +[ + { + "arch": "RNN", + "task": "non_social", + "seed": 0, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 7.440000057220459, + "top_r": 0.03705123940656156 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 1, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 10.039999961853027, + "top_r": 0.09726206921336848 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 2, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 12.199999809265137, + "top_r": 0.09102936182631945 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 3, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 19.8799991607666, + "top_r": 0.1416026760369321 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 4, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 9.880000114440918, + "top_r": 0.06074593045795309 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 5, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 11.760000228881836, + "top_r": 0.06100869691148739 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 6, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 11.760000228881836, + "top_r": 0.047992243964107055 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 7, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 9.720000267028809, + "top_r": 0.040861022303547455 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 8, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 8.359999656677246, + "top_r": 0.046082531832419726 + }, + { + "arch": "RNN", + "task": "non_social", + "seed": 9, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 15.119999885559082, + "top_r": 0.04697112454078715 + }, + { + "arch": "RNN", + "task": "social", + "seed": 0, + "status": "ok", + "n_valid": 16, + "vision": 0.16239520907402039, + "collisions": 60.91999816894531, + "top_r": 0.9606705270007906 + }, + { + "arch": "RNN", + "task": "social", + "seed": 1, + "status": "ok", + "n_valid": 18, + "vision": 0.21668662130832672, + "collisions": 0.800000011920929, + "top_r": 0.6658904391428773 + }, + { + "arch": "RNN", + "task": "social", + "seed": 2, + "status": "ok", + "n_valid": 22, + "vision": 0.3185628652572632, + "collisions": 5.519999980926514, + "top_r": 0.6149421028031533 + }, + { + "arch": "RNN", + "task": "social", + "seed": 3, + "status": "degenerate", + "n_valid": 1, + "vision": 0.24694611132144928, + "collisions": 47.279998779296875, + "top_r": null + }, + { + "arch": "RNN", + "task": "social", + "seed": 4, + "status": "ok", + "n_valid": 23, + "vision": 0.9924950003623962, + "collisions": 150.9199981689453, + "top_r": 0.9469448413117497 + }, + { + "arch": "RNN", + "task": "social", + "seed": 5, + "status": "ok", + "n_valid": 23, + "vision": 0.03936127573251724, + "collisions": 0.5199999809265137, + "top_r": 0.24632233096607034 + }, + { + "arch": "RNN", + "task": "social", + "seed": 6, + "status": "ok", + "n_valid": 25, + "vision": 0.2863073945045471, + "collisions": 1.1200000047683716, + "top_r": 0.6195669473643807 + }, + { + "arch": "RNN", + "task": "social", + "seed": 7, + "status": "degenerate", + "n_valid": 2, + "vision": 0.2092614769935608, + "collisions": 21.84000015258789, + "top_r": null + }, + { + "arch": "RNN", + "task": "social", + "seed": 8, + "status": "ok", + "n_valid": 14, + "vision": 0.13309381902217865, + "collisions": 19.959999084472656, + "top_r": 0.9244864792894123 + }, + { + "arch": "RNN", + "task": "social", + "seed": 9, + "status": "ok", + "n_valid": 24, + "vision": 0.2818363308906555, + "collisions": 5.960000038146973, + "top_r": 0.5286762777215525 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 0, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 8.279999732971191, + "top_r": 0.07776644999751171 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 1, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 8.640000343322754, + "top_r": 0.07086241042694791 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 2, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 9.039999961853027, + "top_r": 0.10571891551983957 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 3, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 10.600000381469727, + "top_r": 0.12438404297121783 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 4, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 13.15999984741211, + "top_r": 0.08272535053372708 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 5, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 4.599999904632568, + "top_r": 0.05911174780800376 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 6, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 6.599999904632568, + "top_r": 0.12048705520440492 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 7, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 15.720000267028809, + "top_r": 0.20280221422494873 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 8, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 13.640000343322754, + "top_r": 0.1197033617321138 + }, + { + "arch": "MLP", + "task": "non_social", + "seed": 9, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 11.920000076293945, + "top_r": 0.06501230641904866 + }, + { + "arch": "MLP", + "task": "social", + "seed": 0, + "status": "ok", + "n_valid": 25, + "vision": 0.981237530708313, + "collisions": 115.08000183105469, + "top_r": 0.68309682518684 + }, + { + "arch": "MLP", + "task": "social", + "seed": 1, + "status": "ok", + "n_valid": 25, + "vision": 0.9638323187828064, + "collisions": 114.72000122070312, + "top_r": 0.6770425529610226 + }, + { + "arch": "MLP", + "task": "social", + "seed": 2, + "status": "ok", + "n_valid": 25, + "vision": 0.9689421057701111, + "collisions": 105.16000366210938, + "top_r": 0.6294873925115205 + }, + { + "arch": "MLP", + "task": "social", + "seed": 3, + "status": "ok", + "n_valid": 25, + "vision": 0.9881038069725037, + "collisions": 236.0800018310547, + "top_r": 0.7859041502514537 + }, + { + "arch": "MLP", + "task": "social", + "seed": 4, + "status": "ok", + "n_valid": 25, + "vision": 0.9528942108154297, + "collisions": 125.0, + "top_r": 0.7282436596506869 + }, + { + "arch": "MLP", + "task": "social", + "seed": 5, + "status": "ok", + "n_valid": 25, + "vision": 0.9758083820343018, + "collisions": 110.36000061035156, + "top_r": 0.6151424539297031 + }, + { + "arch": "MLP", + "task": "social", + "seed": 6, + "status": "ok", + "n_valid": 25, + "vision": 0.9893013834953308, + "collisions": 114.23999786376953, + "top_r": 0.7531067455891379 + }, + { + "arch": "MLP", + "task": "social", + "seed": 7, + "status": "ok", + "n_valid": 25, + "vision": 0.9883432984352112, + "collisions": 141.52000427246094, + "top_r": 0.6580303971938998 + }, + { + "arch": "MLP", + "task": "social", + "seed": 8, + "status": "ok", + "n_valid": 25, + "vision": 0.9682235717773438, + "collisions": 101.19999694824219, + "top_r": 0.6171618473537643 + }, + { + "arch": "MLP", + "task": "social", + "seed": 9, + "status": "ok", + "n_valid": 25, + "vision": 0.9904990196228027, + "collisions": 115.23999786376953, + "top_r": 0.639261743430404 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 0, + "status": "ok", + "n_valid": 24, + "vision": 0.0, + "collisions": 40.2400016784668, + "top_r": 0.28866651056915354 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 1, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 3.4800000190734863, + "top_r": 0.15885310687782797 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 2, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 3.799999952316284, + "top_r": 0.1087223462156361 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 3, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 1.8799999952316284, + "top_r": 0.0647609791042199 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 4, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 34.36000061035156, + "top_r": 0.6156083569490233 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 5, + "status": "degenerate", + "n_valid": 1, + "vision": 0.0, + "collisions": 34.68000030517578, + "top_r": null + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 6, + "status": "ok", + "n_valid": 24, + "vision": 0.0, + "collisions": 3.200000047683716, + "top_r": 0.08990168984095623 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 7, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 121.23999786376953, + "top_r": 0.6383316081199204 + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 8, + "status": "degenerate", + "n_valid": 0, + "vision": 0.0, + "collisions": 46.31999969482422, + "top_r": null + }, + { + "arch": "SSM", + "task": "non_social", + "seed": 9, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 105.68000030517578, + "top_r": 0.5532722247875926 + }, + { + "arch": "SSM", + "task": "social", + "seed": 0, + "status": "ok", + "n_valid": 13, + "vision": 0.9386028051376343, + "collisions": 107.16000366210938, + "top_r": 0.6762547587766922 + }, + { + "arch": "SSM", + "task": "social", + "seed": 1, + "status": "ok", + "n_valid": 19, + "vision": 0.6916567087173462, + "collisions": 41.91999816894531, + "top_r": 0.5728661391129467 + }, + { + "arch": "SSM", + "task": "social", + "seed": 2, + "status": "ok", + "n_valid": 22, + "vision": 0.9957684874534607, + "collisions": 63.91999816894531, + "top_r": 0.5386769468869881 + }, + { + "arch": "SSM", + "task": "social", + "seed": 3, + "status": "degenerate", + "n_valid": 0, + "vision": 0.9954491257667542, + "collisions": 400.9200134277344, + "top_r": null + }, + { + "arch": "SSM", + "task": "social", + "seed": 4, + "status": "ok", + "n_valid": 5, + "vision": 0.4332934021949768, + "collisions": 20.15999984741211, + "top_r": 0.6451686137949718 + }, + { + "arch": "SSM", + "task": "social", + "seed": 5, + "status": "ok", + "n_valid": 25, + "vision": 0.981157660484314, + "collisions": 66.23999786376953, + "top_r": 0.5719388049785209 + }, + { + "arch": "SSM", + "task": "social", + "seed": 6, + "status": "ok", + "n_valid": 24, + "vision": 0.9827544689178467, + "collisions": 63.79999923706055, + "top_r": 0.6399909552314786 + }, + { + "arch": "SSM", + "task": "social", + "seed": 7, + "status": "ok", + "n_valid": 8, + "vision": 0.9846706390380859, + "collisions": 129.55999755859375, + "top_r": 0.5951176365769322 + }, + { + "arch": "SSM", + "task": "social", + "seed": 8, + "status": "ok", + "n_valid": 3, + "vision": 0.8347305655479431, + "collisions": 139.27999877929688, + "top_r": 0.5911175381223174 + }, + { + "arch": "SSM", + "task": "social", + "seed": 9, + "status": "ok", + "n_valid": 8, + "vision": 0.9699001908302307, + "collisions": 279.9599914550781, + "top_r": 0.9463852317330589 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 0, + "status": "ok", + "n_valid": 24, + "vision": 0.0, + "collisions": 18.15999984741211, + "top_r": 0.20624045007674752 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 1, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 177.52000427246094, + "top_r": 0.5466457412672473 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 2, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 247.67999267578125, + "top_r": 0.741521978186391 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 3, + "status": "ok", + "n_valid": 12, + "vision": 0.0, + "collisions": 32.47999954223633, + "top_r": 0.29652097824664553 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 4, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 89.83999633789062, + "top_r": 0.3499635714443735 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 5, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 118.91999816894531, + "top_r": 0.4220079232087302 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 6, + "status": "ok", + "n_valid": 15, + "vision": 0.0, + "collisions": 199.83999633789062, + "top_r": 0.4868188359531619 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 7, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 87.63999938964844, + "top_r": 0.3143182076125383 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 8, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 12.920000076293945, + "top_r": 0.10559527101392678 + }, + { + "arch": "Transformer", + "task": "non_social", + "seed": 9, + "status": "ok", + "n_valid": 25, + "vision": 0.0, + "collisions": 4.920000076293945, + "top_r": 0.0921079793072239 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 0, + "status": "ok", + "n_valid": 23, + "vision": 0.8109381198883057, + "collisions": 66.27999877929688, + "top_r": 0.4658015424447193 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 1, + "status": "ok", + "n_valid": 24, + "vision": 0.816926121711731, + "collisions": 83.04000091552734, + "top_r": 0.37708898276609976 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 2, + "status": "ok", + "n_valid": 14, + "vision": 0.9682235717773438, + "collisions": 81.19999694824219, + "top_r": 0.4850167809908031 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 3, + "status": "ok", + "n_valid": 20, + "vision": 0.8345708847045898, + "collisions": 78.55999755859375, + "top_r": 0.4330689299750323 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 4, + "status": "ok", + "n_valid": 21, + "vision": 0.757125735282898, + "collisions": 76.16000366210938, + "top_r": 0.34081183250426467 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 5, + "status": "ok", + "n_valid": 20, + "vision": 0.6459081768989563, + "collisions": 100.72000122070312, + "top_r": 0.4682438954043613 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 6, + "status": "ok", + "n_valid": 25, + "vision": 0.7647106051445007, + "collisions": 47.84000015258789, + "top_r": 0.37878757497109683 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 7, + "status": "ok", + "n_valid": 24, + "vision": 0.9699800610542297, + "collisions": 95.5999984741211, + "top_r": 0.4278219769902787 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 8, + "status": "ok", + "n_valid": 20, + "vision": 0.7192016243934631, + "collisions": 103.23999786376953, + "top_r": 0.6244838845224036 + }, + { + "arch": "Transformer", + "task": "social", + "seed": 9, + "status": "ok", + "n_valid": 24, + "vision": 0.7394012212753296, + "collisions": 95.95999908447266, + "top_r": 0.3750948684332812 + } +] \ No newline at end of file diff --git a/cross_architecture_study/cross_arch_plsc_residualized.json b/cross_architecture_study/cross_arch_plsc_residualized.json new file mode 100644 index 0000000000000000000000000000000000000000..25c2c291d4439751352808a0c0a2630eceef1c48 --- /dev/null +++ b/cross_architecture_study/cross_arch_plsc_residualized.json @@ -0,0 +1,401 @@ +{ + "RNN": { + "social": [ + [ + 0.9606705270007906, + 156, + 16 + ], + [ + 0.6658904391428773, + 236, + 18 + ], + [ + 0.6149421028031533, + 250, + 22 + ], + [ + 0.9469448413117497, + 256, + 23 + ], + [ + 0.24632233096607034, + 190, + 23 + ], + [ + 0.6195669473643807, + 255, + 25 + ], + [ + 0.9244864792894123, + 212, + 14 + ], + [ + 0.5286762777215525, + 256, + 24 + ] + ], + "non_social": [ + [ + 0.03705123940656156, + 54, + 25 + ], + [ + 0.09726206921336848, + 111, + 25 + ], + [ + 0.09102936182631945, + 108, + 25 + ], + [ + 0.1416026760369321, + 116, + 25 + ], + [ + 0.06074593045795309, + 95, + 25 + ], + [ + 0.06100869691148739, + 122, + 25 + ], + [ + 0.047992243964107055, + 101, + 25 + ], + [ + 0.040861022303547455, + 113, + 25 + ], + [ + 0.046082531832419726, + 120, + 25 + ], + [ + 0.04697112454078715, + 121, + 25 + ] + ] + }, + "MLP": { + "social": [ + [ + 0.68309682518684, + 256, + 25 + ], + [ + 0.6770425529610226, + 256, + 25 + ], + [ + 0.6294873925115205, + 256, + 25 + ], + [ + 0.7859041502514537, + 256, + 25 + ], + [ + 0.7282436596506869, + 256, + 25 + ], + [ + 0.6151424539297031, + 256, + 25 + ], + [ + 0.7531067455891379, + 256, + 25 + ], + [ + 0.6580303971938998, + 256, + 25 + ], + [ + 0.6171618473537643, + 256, + 25 + ], + [ + 0.639261743430404, + 256, + 25 + ] + ], + "non_social": [ + [ + 0.07776644999751171, + 253, + 25 + ], + [ + 0.07086241042694791, + 254, + 25 + ], + [ + 0.10571891551983957, + 255, + 25 + ], + [ + 0.12438404297121783, + 255, + 25 + ], + [ + 0.08272535053372708, + 255, + 25 + ], + [ + 0.05911174780800376, + 255, + 25 + ], + [ + 0.12048705520440492, + 251, + 25 + ], + [ + 0.20280221422494873, + 256, + 25 + ], + [ + 0.1197033617321138, + 231, + 25 + ], + [ + 0.06501230641904866, + 251, + 25 + ] + ] + }, + "SSM": { + "social": [ + [ + 0.6762547587766922, + 253, + 13 + ], + [ + 0.5728661391129467, + 256, + 19 + ], + [ + 0.5386769468869881, + 249, + 22 + ], + [ + 0.6451686137949718, + 249, + 5 + ], + [ + 0.5719388049785209, + 256, + 25 + ], + [ + 0.6399909552314786, + 256, + 24 + ], + [ + 0.5951176365769322, + 239, + 8 + ], + [ + 0.5911175381223174, + 179, + 3 + ], + [ + 0.9463852317330589, + 224, + 8 + ] + ], + "non_social": [ + [ + 0.28866651056915354, + 235, + 24 + ], + [ + 0.15885310687782797, + 137, + 25 + ], + [ + 0.1087223462156361, + 99, + 25 + ], + [ + 0.0647609791042199, + 185, + 25 + ], + [ + 0.6156083569490233, + 244, + 25 + ], + [ + 0.08990168984095623, + 100, + 24 + ], + [ + 0.6383316081199204, + 252, + 25 + ], + [ + 0.5532722247875926, + 235, + 25 + ] + ] + }, + "Transformer": { + "social": [ + [ + 0.4658015424447193, + 251, + 23 + ], + [ + 0.37708898276609976, + 251, + 24 + ], + [ + 0.4850167809908031, + 253, + 14 + ], + [ + 0.4330689299750323, + 254, + 20 + ], + [ + 0.34081183250426467, + 249, + 21 + ], + [ + 0.4682438954043613, + 241, + 20 + ], + [ + 0.37878757497109683, + 250, + 25 + ], + [ + 0.4278219769902787, + 249, + 24 + ], + [ + 0.6244838845224036, + 249, + 20 + ], + [ + 0.3750948684332812, + 248, + 24 + ] + ], + "non_social": [ + [ + 0.20624045007674752, + 20, + 24 + ], + [ + 0.5466457412672473, + 152, + 25 + ], + [ + 0.741521978186391, + 139, + 25 + ], + [ + 0.29652097824664553, + 16, + 12 + ], + [ + 0.3499635714443735, + 91, + 25 + ], + [ + 0.4220079232087302, + 70, + 25 + ], + [ + 0.4868188359531619, + 209, + 15 + ], + [ + 0.3143182076125383, + 181, + 25 + ], + [ + 0.10559527101392678, + 113, + 25 + ], + [ + 0.0921079793072239, + 15, + 25 + ] + ] + } +} \ No newline at end of file diff --git a/cross_architecture_study/figures/cross_arch_cka.png b/cross_architecture_study/figures/cross_arch_cka.png new file mode 100644 index 0000000000000000000000000000000000000000..03f1acaf9e07adb40d3bef8b3f0fc02d0bbdcf91 Binary files /dev/null and b/cross_architecture_study/figures/cross_arch_cka.png differ diff --git a/cross_architecture_study/figures/sharing_vs_vision.png b/cross_architecture_study/figures/sharing_vs_vision.png new file mode 100644 index 0000000000000000000000000000000000000000..25edb5ee309e934de24ea333104bf6cc0a4349d9 Binary files /dev/null and b/cross_architecture_study/figures/sharing_vs_vision.png differ diff --git a/cross_architecture_study/report.md b/cross_architecture_study/report.md new file mode 100644 index 0000000000000000000000000000000000000000..51b450c000f250f6815106032c85e8107cb54b9f --- /dev/null +++ b/cross_architecture_study/report.md @@ -0,0 +1,215 @@ +# Cross-Architecture Representation Study + +**Question.** Do different agent architectures, trained identically on the +Zhang et al. (2025) chaser–explorer task, learn the *same* internal +representations of the social interaction? Or does the phenomenon the paper +reports — genuine shared neural dynamics between two independently trained +agents — depend on the architecture? + +**Design.** Four architectures, identical in everything except the memory +mechanism: reward structure, 10×10 world, 100-step episodes, seeds 0–9, +analysis dimension 256, PPO settings, evaluation protocol, and analysis +pipeline. They span a memory ladder: + +| architecture | memory mechanism | internal state at step t | +|---|---|---| +| MLP (`k=8` frame stack) | none (finite window) | pure function of the last 8 observations | +| SSM (gated diagonal LRU) | compressed, **linear** dynamics | eigenvalue-controlled linear state | +| RNN (paper's ReLU RNN) | compressed, **nonlinear** dynamics | chaotic nonlinear recurrence | +| Transformer (2-layer causal) | none evolving; re-attends history | pre-head residual recomputed per step | + +Each was trained as a full 20-pair experiment (`mouse-run-run-1` for the RNN, +`mouse-run-run-2-{mlp,ssm,transformer}` for the rest): 20,000 PPO updates × +40 episodes × 100 steps per pair, **20/20 units completed on the first +attempt** for every architecture, zero failed/interrupted attempts. + +--- + +## 1. Behavior: the architectures converged to different strategies + +Chaser capability against a standardized random explorer (100 episodes × +100 steps), and the explorer's learned escape (collisions it suffers against a +random chaser — lower is better escape): + +| architecture | chaser collisions ↑ | chaser partner-in-vision | explorer collisions-suffered ↓ | +|---|---:|---:|---:| +| RNN | 8.98 ± 5.70 | 0.61 | 0.57 | +| MLP | 26.93 ± 4.00 | 0.93 | 0.92 | +| SSM | 35.92 ± 3.56 | 0.94 | 0.45 | +| Transformer | 41.31 ± 3.29 | 0.96 | 0.28 | + +All four learned role-specific social behavior (non-social controls collide +2.7–3.4× regardless). But the **RNN converged to a qualitatively different +strategy**: a low-vision "stealth" chaser (61% partner-in-vision, 9 collisions) +that tracks the explorer from memory, whereas MLP/SSM/Transformer converged to +high-vision "pursuit" chasers (93–96% vision, 27–41 collisions) that stay +glued to the explorer. This behavioral fork is the first sign the architectures +are not learning the same thing — and it turns out to drive the representation +difference. + +The raw capability ranking (Transformer > SSM > MLP > RNN) is **not** the +headline. A chaser that collides 41×/episode by never leaving the explorer's +side is behaviorally strong but representationally shallow. The scientific +question is about the *internal representation*, below. + +## 2. Event decoding replicates across architectures (C2) + +Linear-classifier balanced accuracy for social events decoded from single-agent +hidden states (5-fold CV, shuffled controls ≈ 0.50 throughout): + +| architecture | collision | partner escape | partner approach | +|---|---:|---:|---:| +| RNN | 0.983 | 0.888 | 0.886 | +| MLP | 0.914 | 0.780 | 0.778 | +| SSM | 0.850 | 0.768 | 0.733 | +| Transformer | 0.847 | 0.761 | 0.759 | + +Every architecture encodes collision and partner approach/escape above chance +in social agents — the paper's C2 claim is **architecture-robust**. Decoding +accuracy tracks the memory ladder (RNN > MLP ≈ SSM > Transformer), but the +qualitative result (social events are linearly represented) holds for all. + +## 3. Shared dimensions: a methodological correction, then the key result (C3) + +The paper's headline is C3: social agents develop *shared* neural dimensions +across the two independently trained networks (PLSC). Measuring this across +architectures exposed a confound that the analysis had to fix. + +**The raw cross-agent correlation is contaminated by time-locked scaffolds.** +The Transformer's `top_dim_correlation` was **0.957 in the non-social control** +— higher than its social value, and higher than any architecture's social +score. There is no partner input in the non-social task, so this cannot be +genuine sharing. The cause: the Transformer's positional embedding makes every +timestep a deterministic function of absolute position t; both agents' networks +carry this identical positional scaffold, and it dominates the top singular +dimension. The RNN and MLP have no such clean scaffold (their "clocks" are noisy +or absent), so their raw non-social correlation stays near zero — which is +exactly why the raw statistic *looked* fine until the Transformer exposed it. + +**Fix (standard in neuroscience): remove the evoked/time-locked component.** +Subtract the across-episode mean at each timestep — the response identical +across episodes at a given t (positional embeddings, hidden-state clocks, mean +approach dynamics) — and run PLSC on the episode-specific residual, where +genuine interaction lives. This removes every architecture's deterministic +temporal scaffold uniformly, making the effect size comparable. + +**Aggregate residualized top-dim correlation (interaction-driven sharing):** + +| architecture | social | non-social | separation | +|---|---:|---:|---:| +| RNN | 0.688 ± 0.23 | 0.067 | **0.62** | +| MLP (floor) | 0.679 ± 0.06 | 0.103 | 0.58 | +| SSM | 0.642 ± 0.12 | 0.315 | 0.33 | +| Transformer | 0.438 ± 0.08 | 0.356 | **0.08** | + +The MLP is the **common-input floor**: being memoryless, any cross-agent sharing +it shows is by construction reactive (both nets responding to the same visible +game state). That its social sharing (0.679) nearly matches the RNN's aggregate +(0.688) is the crux — the aggregate mean *hides* the real difference. The +Transformer's social/non-social separation collapses to 0.08 once its positional +scaffold is removed: **its apparent sharing was mostly the scaffold**. + +**The discriminating view: sharing vs behavioral coupling (vision).** The MLP +can only share when both agents see each other. So the question is whether an +architecture achieves shared representation *at low mutual vision* — which +requires genuine internal (memory/prediction) coupling, not common input. +Residualized social `top_r`, split by chaser partner-in-vision: + +| architecture | low vision (<40%) | high vision (≥40%) | +|---|---:|---:| +| **RNN** | **0.652 (n=7)** | 0.947 (n=1) | +| MLP | — (0 units) | 0.679 (n=10) | +| SSM | — (0 units) | 0.642 (n=9) | +| Transformer | — (0 units) | 0.438 (n=10) | + +This is the result. **Only the RNN develops shared representation without the +agents seeing each other** — 7 of its 8 analyzable social units sit below 40% +mutual vision yet reach 0.65 residualized correlation. Every social unit of the +other three architectures lives at high vision; their sharing is vision-gated, +i.e. common-input-like. MLP *cannot* do otherwise (no memory). SSM and +Transformer *have* memory mechanisms but did not use them to build low-vision +coordination — they converged to the reactive, high-vision solution. + +## 4. Representational geometry: the RNN is the outlier (CKA) + +Feeding the **same** game-state observation stream through every architecture's +social chaser and comparing representations with linear CKA (rotation/scaling +invariant): + +| | RNN | MLP | SSM | Transformer | +|---|---:|---:|---:|---:| +| RNN | 1.00 | 0.68 | 0.64 | 0.67 | +| MLP | 0.68 | 1.00 | 0.71 | 0.78 | +| SSM | 0.64 | 0.71 | 1.00 | 0.76 | +| Transformer | 0.67 | 0.78 | 0.76 | 1.00 | + +The **RNN is the representational outlier** (0.64–0.68 with everything else), +while MLP/SSM/Transformer cluster together (0.71–0.78). A third independent +method agrees with behavior and PLSC: the RNN encodes the task differently; the +other three converge to a more mutually similar (reactive) geometry. + +## 5. Degeneracy is itself an architecture property + +The paper excludes "degenerate" episodes where agents get stuck. The +**SSM is markedly more prone to this**: 1/10 social and 2/10 non-social units +were fully degenerate (too few valid episodes to analyze), and several more had +only 3–8 of 25 episodes valid — versus **zero fully-degenerate units for RNN, +MLP, and Transformer**. The linear recurrence appears to collapse toward +fixed-point / stuck behavior more readily — a concrete failure mode of the +linear-dynamics inductive bias on this task. + +--- + +## Answer to the question + +**Different architectures do not learn the same internal representations.** +Three independent methods converge: + +1. **Behavior** — the RNN alone converges to a low-vision, memory-based stealth + strategy; the others to high-vision reactive pursuit. +2. **Shared dimensions (residualized, vision-conditioned)** — the RNN alone + develops genuine interaction-driven sharing when the agents *cannot see each + other*; the others' sharing is vision-gated and, for the Transformer, mostly + a positional scaffold once corrected. +3. **CKA** — the RNN is the representational outlier; the other three cluster. + +The paper's hallmark phenomenon — genuine internal shared dynamics between +independently trained agents — is **not architecture-universal**. It emerged +cleanly only in the vanilla RNN. The memoryless MLP cannot produce it (its +sharing is the common-input floor). The SSM and Transformer, despite having the +capacity for memory, did not: their inductive biases steered them to a reactive, +high-vision solution whose apparent sharing is largely explained by behavioral +and input coupling, plus (Transformer) a deterministic temporal scaffold. + +## Honest confounds and limitations + +- **Strategy mediates representation.** The architectures converged to different + behavioral strategies, so the representation difference is partly *mediated* by + strategy, not a direct architecture→representation effect. But that is itself a + finding: the architecture's inductive bias shapes *which* solution is found, + and only the RNN's solution has genuine internal shared dynamics. Isolating a + direct effect would require conditioning on matched behavior (e.g. training all + architectures to the same vision regime), which these runs do not do. +- **Not parameter-matched.** RNN 119k, SSM 382k, MLP 477k, Transformer 1.24M + params (analysis dim held at 256). The RNN's distinctness is not a + capacity-advantage artifact — it is the *smallest* model. +- **Time-mean subtraction** removes any genuinely time-locked coordination along + with the scaffold; for random-initial-position episodes the interaction timing + is episode-specific, so this should be conservative, but it is an assumption. +- **PLSC significant-dimension counts saturate** at these sample sizes (MLP hits + 256/256 in both conditions); the effect size (residualized top correlation), + not the count, is the discriminating statistic and is what this report uses. +- The **perturbation (C5)** and neural-action-space (C4) analyses are not yet run + across architectures; C5 in particular (does removing shared dimensions impair + social behavior) would test causality of the RNN's shared dimensions. + +## Artifacts + +- Per-experiment behavior/eval/rollouts/reports: `runs/{experiment}/`, + `runs/reports/{experiment}/`. +- Neural analyses: `runs/analyses/{experiment}/neural_summary.json`. +- Cross-architecture: `runs/analyses/cross_arch_plsc_residualized.json`, + `cross_arch_perunit.json`, `cross_arch_cka.json`. +- Figures: `runs/analyses/figures/{sharing_vs_vision,cross_arch_cka}.png`. +- Analysis code: `mouse_run_run/analysis.py`, + `scripts/analysis/neural_analyses.py`. diff --git a/derived/reports/MANIFEST.json b/derived/reports/MANIFEST.json new file mode 100644 index 0000000000000000000000000000000000000000..6077a957c80a14e90c599cfcfa671f215e380a5c --- /dev/null +++ b/derived/reports/MANIFEST.json @@ -0,0 +1,71 @@ +{ + "analysis_script": "/root/mouse-run-run-perf/scripts/analysis/summarize_experiment.py", + "generated_at": "2026-07-02T11:51:02.815710+00:00", + "input_hashes": { + "evaluations.jsonl": "265ed5550a98149da5384ac6cbcd551db5b121176f7046a4ca58d89531a77cc7", + "exclusions.jsonl": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "rollouts.jsonl": "140147549967098f1a78dadab8e5888eb41a37389c094f79b623d22d7f0bd74a", + "runs.jsonl": "bfab762e5dd3673a8e08fa06635ebebd07bc02783f016fc26e56fee41756e2d8" + }, + "input_tables": { + "evaluations.jsonl": "runs/tables/mouse-run-run-2-mlp/evaluations.jsonl", + "exclusions.jsonl": "runs/tables/mouse-run-run-2-mlp/exclusions.jsonl", + "rollouts.jsonl": "runs/tables/mouse-run-run-2-mlp/rollouts.jsonl", + "runs.jsonl": "runs/tables/mouse-run-run-2-mlp/runs.jsonl" + }, + "output_hashes": { + "evaluation_comparison": "f884d82cdc4bac267de1ab48120986784960a6448d8c7f4502cc3b48c4369735", + "training_curves": "f79e51dd791199e40a8057983a80cb9405bc95f186a2e5ee61eae94544f41bac" + }, + "outputs": { + "figures": [ + "runs/reports/mouse-run-run-2-mlp/figures/training_curves.png", + "runs/reports/mouse-run-run-2-mlp/figures/evaluation_comparison.png" + ], + "report": "runs/reports/mouse-run-run-2-mlp/report.md", + "summary": "runs/reports/mouse-run-run-2-mlp/summary.json" + }, + "provenance": { + "command": "scripts/analysis/summarize_experiment.py runs/tables/mouse-run-run-2-mlp", + "created_at": "2026-07-02T11:51:02.816299+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "schema_version": 2 +} diff --git a/derived/reports/figures/evaluation_comparison.png b/derived/reports/figures/evaluation_comparison.png new file mode 100644 index 0000000000000000000000000000000000000000..c02305c2d9ef6052eb0b0e3ddf84fc224df27e56 --- /dev/null +++ b/derived/reports/figures/evaluation_comparison.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f884d82cdc4bac267de1ab48120986784960a6448d8c7f4502cc3b48c4369735 +size 172530 diff --git a/derived/reports/figures/training_curves.png b/derived/reports/figures/training_curves.png new file mode 100644 index 0000000000000000000000000000000000000000..766bd8cde2ba58dcbe0f6fe4fbb7998080405c97 --- /dev/null +++ b/derived/reports/figures/training_curves.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f79e51dd791199e40a8057983a80cb9405bc95f186a2e5ee61eae94544f41bac +size 386863 diff --git a/derived/reports/report.md b/derived/reports/report.md new file mode 100644 index 0000000000000000000000000000000000000000..39d17c4116068f948f561dd2f56a164524a47497 --- /dev/null +++ b/derived/reports/report.md @@ -0,0 +1,88 @@ +# MARL Experiment Summary + +Generated: `2026-07-02T11:51:02.813569+00:00` +Tables: `runs/tables/mouse-run-run-2-mlp` + +## Valid Pairs + +| Task | Valid pairs | +| --- | ---: | +| `non_social` | 10 | +| `social` | 10 | + +## Training Metrics (final update, mean ± std across pairs) + +| Metric | `non_social` (n=10) | `social` (n=10) | +| --- | ---: | ---: | +| collisions per episode | 1.69 ± 0.547 | 20.353 ± 1.679 | +| chaser return | 6.472 ± 0.343 | 21.552 ± 1.797 | +| explorer return | 56.474 ± 10.32 | 21.788 ± 4.747 | +| chaser partner vision | 0 ± 0 | 0.936 ± 0.013 | +| explorer partner vision | 0 ± 0 | 0.936 ± 0.013 | +| chaser new fields | 82.878 ± 1.721 | 49.655 ± 2.498 | +| explorer new fields | 71.42 ± 6.901 | 58.828 ± 3.202 | +| final distance | 4.834 ± 0.443 | 1.67 ± 0.112 | +| policy loss | -0.000661 ± 0.000258 | -0.004 ± 0.000856 | +| value loss | 3.652 ± 0.263 | 4.878 ± 0.277 | +| entropy | 0.057 ± 0.031 | 0.151 ± 0.02 | +| approx kl | 0.000278 ± 0.000228 | 0.00088 ± 0.000945 | +| chaser grad norm | 0.078 ± 0.023 | 0.41 ± 0.212 | +| explorer grad norm | 0.584 ± 0.277 | 0.549 ± 0.25 | +| max param abs | 1.452 ± 0.159 | 1.278 ± 0.164 | + +## Random-Opponent Evaluation + +### Opponent: `random_chaser` + +| Metric | `non_social` (n=10) | `social` (n=10) | +| --- | ---: | ---: | +| collisions per episode | 2.307 ± 0.275 | 0.924 ± 0.308 | +| chaser return | -3.304 ± 0.037 | -2.234 ± 0.333 | +| explorer return | 55.076 ± 9.319 | 27.815 ± 6.397 | +| chaser partner vision | 0 ± 0 | 0.241 ± 0.028 | +| explorer partner vision | 0 ± 0 | 0.241 ± 0.028 | +| chaser new fields | 33.775 ± 0.186 | 33.883 ± 0.057 | +| explorer new fields | 70.527 ± 6.243 | 52.399 ± 4.267 | +| final distance | 5.038 ± 0.312 | 5.773 ± 0.251 | +| average distance | 5.106 ± 0.068 | 5.833 ± 0.193 | +| degenerate fraction | 0.888 ± 0.031 | 0.892 ± 0.013 | + +### Opponent: `random_explorer` + +| Metric | `non_social` (n=10) | `social` (n=10) | +| --- | ---: | ---: | +| collisions per episode | 2.681 ± 0.243 | 26.935 ± 4.003 | +| chaser return | 5.923 ± 0.544 | 25.334 ± 4.144 | +| explorer return | -1.283 ± 0.15 | -28.943 ± 4.167 | +| chaser partner vision | 0 ± 0 | 0.932 ± 0.017 | +| explorer partner vision | 0 ± 0 | 0.932 ± 0.017 | +| chaser new fields | 80.372 ± 2.756 | 31.427 ± 1.323 | +| explorer new fields | 32.887 ± 0.084 | 26.529 ± 1.279 | +| final distance | 4.955 ± 0.275 | 1.715 ± 0.099 | +| average distance | 5.141 ± 0.061 | 1.921 ± 0.113 | +| degenerate fraction | 0.845 ± 0.014 | 0.95 ± 0.02 | + +Failed evaluations: 0 + +## Figures + +### training curves + +![training_curves](figures/training_curves.png) + +Training curves, mean ± std across 20 valid pairs. + +### evaluation comparison + +![evaluation_comparison](figures/evaluation_comparison.png) + +Evaluation metrics over 40 successful evaluations, grouped by opponent mode and task. + +## Degenerate Episode Exclusions + +| Excluded episodes | Episodes with records | Records | +| ---: | ---: | ---: | +| 0 | 0 | 0 | + +## Failed Attempts +- count: 0 diff --git a/derived/reports/summary.json b/derived/reports/summary.json new file mode 100644 index 0000000000000000000000000000000000000000..89c8cca1c7610dcadaecd6a8c5a00ce6adf671ec --- /dev/null +++ b/derived/reports/summary.json @@ -0,0 +1,544 @@ +{ + "evaluation_metrics": { + "failed_count": 0, + "non_social": { + "random_chaser": { + "average_distance": { + "max": 5.199133396148682, + "mean": 5.105685377120972, + "min": 4.98706579208374, + "n": 10, + "std": 0.06848503924382211 + }, + "chaser_new_fields": { + "max": 33.97999954223633, + "mean": 33.77499961853027, + "min": 33.43000030517578, + "n": 10, + "std": 0.18618548859771447 + }, + "chaser_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "chaser_return": { + "max": -3.2539987564086914, + "mean": -3.3035988092422484, + "min": -3.3739986419677734, + "n": 10, + "std": 0.037113861963604215 + }, + "collisions_per_episode": { + "max": 2.919999837875366, + "mean": 2.306999957561493, + "min": 1.8600000143051147, + "n": 10, + "std": 0.27513812098917756 + }, + "degenerate_fraction": { + "max": 0.97, + "mean": 0.8880000000000001, + "min": 0.85, + "n": 10, + "std": 0.030594117081556706 + }, + "explorer_new_fields": { + "max": 80.68999481201172, + "mean": 70.52699851989746, + "min": 60.68000030517578, + "n": 10, + "std": 6.243129016079577 + }, + "explorer_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "explorer_return": { + "max": 70.2699966430664, + "mean": 55.076498413085936, + "min": 40.404998779296875, + "n": 10, + "std": 9.31878213892601 + }, + "final_distance": { + "max": 5.588272571563721, + "mean": 5.037810468673706, + "min": 4.463276386260986, + "n": 10, + "std": 0.311963295971499 + }, + "n": 10 + }, + "random_explorer": { + "average_distance": { + "max": 5.262848854064941, + "mean": 5.140882539749145, + "min": 5.038296699523926, + "n": 10, + "std": 0.0614441857456308 + }, + "chaser_new_fields": { + "max": 86.43999481201172, + "mean": 80.3719985961914, + "min": 76.8699951171875, + "n": 10, + "std": 2.755952134734959 + }, + "chaser_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "chaser_return": { + "max": 7.12199592590332, + "mean": 5.922596454620361, + "min": 5.221996784210205, + "n": 10, + "std": 0.5440438394983717 + }, + "collisions_per_episode": { + "max": 3.049999952316284, + "mean": 2.68099992275238, + "min": 2.309999942779541, + "n": 10, + "std": 0.24341115086841367 + }, + "degenerate_fraction": { + "max": 0.87, + "mean": 0.845, + "min": 0.83, + "n": 10, + "std": 0.013601470508735456 + }, + "explorer_new_fields": { + "max": 33.040000915527344, + "mean": 32.88699913024902, + "min": 32.790000915527344, + "n": 10, + "std": 0.08391110303705251 + }, + "explorer_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "explorer_return": { + "max": -0.9799999594688416, + "mean": -1.282999974489212, + "min": -1.5199999809265137, + "n": 10, + "std": 0.14951924109474032 + }, + "final_distance": { + "max": 5.392402648925781, + "mean": 4.9553193092346195, + "min": 4.301638603210449, + "n": 10, + "std": 0.2753013389573719 + }, + "n": 10 + } + }, + "social": { + "random_chaser": { + "average_distance": { + "max": 6.055323123931885, + "mean": 5.832867050170899, + "min": 5.418818473815918, + "n": 10, + "std": 0.1926549744871535 + }, + "chaser_new_fields": { + "max": 33.95000076293945, + "mean": 33.88299980163574, + "min": 33.73999786376953, + "n": 10, + "std": 0.05728056247021595 + }, + "chaser_partner_vision": { + "max": 0.2913999855518341, + "mean": 0.24062999188899994, + "min": 0.204599991440773, + "n": 10, + "std": 0.028305234515883754 + }, + "chaser_return": { + "max": -1.4439992904663086, + "mean": -2.2341991543769835, + "min": -2.5089991092681885, + "n": 10, + "std": 0.3334365212524313 + }, + "collisions_per_episode": { + "max": 1.659999966621399, + "mean": 0.9239999949932098, + "min": 0.6699999570846558, + "n": 10, + "std": 0.3079350544139646 + }, + "degenerate_fraction": { + "max": 0.91, + "mean": 0.892, + "min": 0.87, + "n": 10, + "std": 0.013266499161421611 + }, + "explorer_new_fields": { + "max": 57.55999755859375, + "mean": 52.3989990234375, + "min": 43.37999725341797, + "n": 10, + "std": 4.266789586753233 + }, + "explorer_partner_vision": { + "max": 0.2913999855518341, + "mean": 0.24062999188899994, + "min": 0.204599991440773, + "n": 10, + "std": 0.028305234515883754 + }, + "explorer_return": { + "max": 35.744998931884766, + "mean": 27.815499019622802, + "min": 14.429999351501465, + "n": 10, + "std": 6.396563521591426 + }, + "final_distance": { + "max": 6.047131061553955, + "mean": 5.773358011245728, + "min": 5.25391149520874, + "n": 10, + "std": 0.2512269355601858 + }, + "n": 10 + }, + "random_explorer": { + "average_distance": { + "max": 2.13496994972229, + "mean": 1.921411967277527, + "min": 1.7290464639663696, + "n": 10, + "std": 0.11278872658258984 + }, + "chaser_new_fields": { + "max": 33.89999771118164, + "mean": 31.42699909210205, + "min": 28.920000076293945, + "n": 10, + "std": 1.3234575205271881 + }, + "chaser_partner_vision": { + "max": 0.9477999806404114, + "mean": 0.9316199898719788, + "min": 0.8927000164985657, + "n": 10, + "std": 0.017225497745840646 + }, + "chaser_return": { + "max": 36.003997802734375, + "mean": 25.334495544433594, + "min": 20.284996032714844, + "n": 10, + "std": 4.144235438399579 + }, + "collisions_per_episode": { + "max": 37.13999938964844, + "mean": 26.934999465942383, + "min": 21.850000381469727, + "n": 10, + "std": 4.003446545522908 + }, + "degenerate_fraction": { + "max": 0.98, + "mean": 0.95, + "min": 0.92, + "n": 10, + "std": 0.020493901531919184 + }, + "explorer_new_fields": { + "max": 28.149999618530273, + "mean": 26.528999519348144, + "min": 23.78999900817871, + "n": 10, + "std": 1.2791048421652784 + }, + "explorer_partner_vision": { + "max": 0.9477999806404114, + "mean": 0.9316199898719788, + "min": 0.8927000164985657, + "n": 10, + "std": 0.017225497745840646 + }, + "explorer_return": { + "max": -23.260000228881836, + "mean": -28.943499755859374, + "min": -39.125, + "n": 10, + "std": 4.167097957623246 + }, + "final_distance": { + "max": 1.8503373861312866, + "mean": 1.7153441309928894, + "min": 1.5262326002120972, + "n": 10, + "std": 0.09933532411070188 + }, + "n": 10 + } + } + }, + "failed_attempts": [], + "figures": [ + { + "caption": "Training curves, mean \u00b1 std across 20 valid pairs.", + "name": "training_curves", + "path": "runs/reports/mouse-run-run-2-mlp/figures/training_curves.png" + }, + { + "caption": "Evaluation metrics over 40 successful evaluations, grouped by opponent mode and task.", + "name": "evaluation_comparison", + "path": "runs/reports/mouse-run-run-2-mlp/figures/evaluation_comparison.png" + } + ], + "generated_at": "2026-07-02T11:51:02.813569+00:00", + "rollout_exclusions": { + "episodes_with_exclusion_records": 0, + "excluded_degenerate_episodes": 0, + "records": 0 + }, + "schema_version": 2, + "tables_root": "runs/tables/mouse-run-run-2-mlp", + "training_metrics": { + "non_social": { + "approx_kl": { + "max": 0.0007710583850932262, + "mean": 0.00027836359963407043, + "min": 3.927184380664306e-06, + "n": 10, + "std": 0.00022781666240450588 + }, + "chaser_grad_norm": { + "max": 0.11076641827821732, + "mean": 0.07840223964303732, + "min": 0.04017048981040716, + "n": 10, + "std": 0.023169039813528615 + }, + "chaser_new_fields": { + "max": 87.07500457763672, + "mean": 82.87750015258788, + "min": 80.4749984741211, + "n": 10, + "std": 1.7213541677267494 + }, + "chaser_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "chaser_return": { + "max": 7.300001621246338, + "mean": 6.471501016616822, + "min": 6.015000820159912, + "n": 10, + "std": 0.34321339840677706 + }, + "collisions_per_episode": { + "max": 2.7250001430511475, + "mean": 1.690000021457672, + "min": 0.9750000238418579, + "n": 10, + "std": 0.5471745917052043 + }, + "entropy": { + "max": 0.13418130949139595, + "mean": 0.05714834872633219, + "min": 0.02094564540311694, + "n": 10, + "std": 0.030692080850040007 + }, + "explorer_grad_norm": { + "max": 1.1887003779411316, + "mean": 0.583899623900652, + "min": 0.2844117097556591, + "n": 10, + "std": 0.27718636321364126 + }, + "explorer_new_fields": { + "max": 82.75, + "mean": 71.42000007629395, + "min": 60.57500076293945, + "n": 10, + "std": 6.900640552089833 + }, + "explorer_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "explorer_return": { + "max": 73.30000305175781, + "mean": 56.473750305175784, + "min": 40.07500076293945, + "n": 10, + "std": 10.319657580882787 + }, + "final_distance": { + "max": 5.683727741241455, + "mean": 4.834419059753418, + "min": 4.144174098968506, + "n": 10, + "std": 0.4431673642962756 + }, + "max_param_abs": { + "max": 1.7384926080703735, + "mean": 1.452017331123352, + "min": 1.2401388883590698, + "n": 10, + "std": 0.15908309762880601 + }, + "n": 10, + "policy_loss": { + "max": -0.0002411644566997495, + "mean": -0.000661178331059098, + "min": -0.0011177757633866747, + "n": 10, + "std": 0.00025768936399840804 + }, + "value_loss": { + "max": 4.014994144439697, + "mean": 3.651918852329254, + "min": 3.322889506816864, + "n": 10, + "std": 0.2633674319523318 + } + }, + "social": { + "approx_kl": { + "max": 0.002805363851033249, + "mean": 0.00087992062623089, + "min": -0.0002899035942244277, + "n": 10, + "std": 0.0009453947863305684 + }, + "chaser_grad_norm": { + "max": 0.9264960587024689, + "mean": 0.41003721952438354, + "min": 0.188689436763525, + "n": 10, + "std": 0.2119915449421527 + }, + "chaser_new_fields": { + "max": 52.92499923706055, + "mean": 49.65500030517578, + "min": 45.57500076293945, + "n": 10, + "std": 2.4984694774842433 + }, + "chaser_partner_vision": { + "max": 0.9552500247955322, + "mean": 0.9360250353813171, + "min": 0.9205000400543213, + "n": 10, + "std": 0.012655263263627304 + }, + "chaser_return": { + "max": 24.285001754760742, + "mean": 21.551751327514648, + "min": 18.48750114440918, + "n": 10, + "std": 1.796639770827262 + }, + "collisions_per_episode": { + "max": 23.200000762939453, + "mean": 20.35250053405762, + "min": 17.625, + "n": 10, + "std": 1.6785577964252976 + }, + "entropy": { + "max": 0.19220001995563507, + "mean": 0.15101509019732476, + "min": 0.12773804366588593, + "n": 10, + "std": 0.020022485274942464 + }, + "explorer_grad_norm": { + "max": 1.1695344895124435, + "mean": 0.5494304157793521, + "min": 0.23243309557437897, + "n": 10, + "std": 0.2504391396265317 + }, + "explorer_new_fields": { + "max": 65.55000305175781, + "mean": 58.827500915527345, + "min": 52.775001525878906, + "n": 10, + "std": 3.201532040899581 + }, + "explorer_partner_vision": { + "max": 0.9552500247955322, + "mean": 0.9360250353813171, + "min": 0.9205000400543213, + "n": 10, + "std": 0.012655263263627304 + }, + "explorer_return": { + "max": 31.037500381469727, + "mean": 21.787500286102294, + "min": 12.71250057220459, + "n": 10, + "std": 4.746788332300563 + }, + "final_distance": { + "max": 1.8305482864379883, + "mean": 1.670472300052643, + "min": 1.4895528554916382, + "n": 10, + "std": 0.1115461884593616 + }, + "max_param_abs": { + "max": 1.6354095935821533, + "mean": 1.2778974533081056, + "min": 1.0662928819656372, + "n": 10, + "std": 0.16398531671135025 + }, + "n": 10, + "policy_loss": { + "max": -0.0027591757117768623, + "mean": -0.00390199398393738, + "min": -0.005259661337355226, + "n": 10, + "std": 0.0008555915976877569 + }, + "value_loss": { + "max": 5.285007953643799, + "mean": 4.877697789669037, + "min": 4.4600828886032104, + "n": 10, + "std": 0.27699674303611843 + } + } + }, + "valid_pairs_by_task": { + "non_social": 10, + "social": 10 + } +} diff --git a/derived/tables/MANIFEST.json b/derived/tables/MANIFEST.json new file mode 100644 index 0000000000000000000000000000000000000000..e3e90f1807c65594f9f7d0d7d7d768972203fbe6 --- /dev/null +++ b/derived/tables/MANIFEST.json @@ -0,0 +1,67 @@ +{ + "experiment_root": "runs/mouse-run-run-2-mlp", + "generated_at": "2026-07-02T11:51:00.748915+00:00", + "output_hashes": { + "evaluations": "265ed5550a98149da5384ac6cbcd551db5b121176f7046a4ca58d89531a77cc7", + "exclusions": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "rollouts": "140147549967098f1a78dadab8e5888eb41a37389c094f79b623d22d7f0bd74a", + "runs": "bfab762e5dd3673a8e08fa06635ebebd07bc02783f016fc26e56fee41756e2d8" + }, + "outputs": { + "evaluations": "runs/tables/mouse-run-run-2-mlp/evaluations.jsonl", + "exclusions": "runs/tables/mouse-run-run-2-mlp/exclusions.jsonl", + "rollouts": "runs/tables/mouse-run-run-2-mlp/rollouts.jsonl", + "runs": "runs/tables/mouse-run-run-2-mlp/runs.jsonl" + }, + "panel_path": "runs/tables/mouse-run-run-2-mlp/panels.json", + "provenance": { + "command": "scripts/transform/build_tables.py runs/mouse-run-run-2-mlp", + "created_at": "2026-07-02T11:51:00.749193+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "row_counts": { + "evaluations": 40, + "exclusions": 0, + "rollouts": 20, + "runs": 20 + }, + "schema_version": 1, + "transform_script": "/root/mouse-run-run-perf/scripts/transform/build_tables.py" +} diff --git a/derived/tables/evaluations.jsonl b/derived/tables/evaluations.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..8c188488f03f9365f823d7f95592466684a42c9d --- /dev/null +++ b/derived/tables/evaluations.jsonl @@ -0,0 +1,40 @@ +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.620987+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.200304985046387, "chaser_new_fields": 80.66999816894531, "chaser_partner_vision": 0.0, "chaser_return": 5.999996185302734, "chaser_stuck_run_steps": 0.019999999552965164, "collisions_per_episode": 2.3999998569488525, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 32.79999923706055, "explorer_partner_vision": 0.0, "explorer_return": -1.3700000047683716, "explorer_stuck_run_steps": 2.669999837875366, "final_distance": 5.008528709411621, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 1, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.718185+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.077332973480225, "chaser_new_fields": 33.65999984741211, "chaser_partner_vision": 0.0, "chaser_return": -3.331998825073242, "chaser_stuck_run_steps": 2.700000047683716, "collisions_per_episode": 2.2699999809265137, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 74.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 61.10499954223633, "explorer_stuck_run_steps": 0.07999999821186066, "final_distance": 5.209523677825928, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 2, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.818944+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.202634334564209, "chaser_new_fields": 78.47000122070312, "chaser_partner_vision": 0.0, "chaser_return": 5.575997352600098, "chaser_stuck_run_steps": 0.019999999552965164, "collisions_per_episode": 2.309999942779541, "degenerate_episodes": 83, "degenerate_fraction": 0.83, "episodes": 100, "explorer_new_fields": 32.97999954223633, "explorer_partner_vision": 0.0, "explorer_return": -1.0699999332427979, "explorer_stuck_run_steps": 2.7100000381469727, "final_distance": 5.011303424835205, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 3, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.915491+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.171538352966309, "chaser_new_fields": 33.95000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.2539987564086914, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 2.0899999141693115, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 71.5, "explorer_partner_vision": 0.0, "explorer_return": 56.59000015258789, "explorer_stuck_run_steps": 0.3100000023841858, "final_distance": 4.9595184326171875, "same_state_run_steps": 0.05999999865889549}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 4, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.015998+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.099634170532227, "chaser_new_fields": 81.38999938964844, "chaser_partner_vision": 0.0, "chaser_return": 6.113996505737305, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.049999952316284, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 32.849998474121094, "explorer_partner_vision": 0.0, "explorer_return": -1.3849999904632568, "explorer_stuck_run_steps": 2.700000047683716, "final_distance": 4.918816566467285, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 5, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.112222+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.199133396148682, "chaser_new_fields": 33.43000030517578, "chaser_partner_vision": 0.0, "chaser_return": -3.3739986419677734, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 2.049999952316284, "degenerate_episodes": 85, "degenerate_fraction": 0.85, "episodes": 100, "explorer_new_fields": 80.68999481201172, "explorer_partner_vision": 0.0, "explorer_return": 70.2699966430664, "explorer_stuck_run_steps": 0.7799999713897705, "final_distance": 4.922343730926514, "same_state_run_steps": 0.08999999612569809}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 6, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.212404+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.134275913238525, "chaser_new_fields": 86.43999481201172, "chaser_partner_vision": 0.0, "chaser_return": 7.12199592590332, "chaser_stuck_run_steps": 0.019999999552965164, "collisions_per_episode": 2.6399998664855957, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 32.790000915527344, "explorer_partner_vision": 0.0, "explorer_return": -1.2799999713897705, "explorer_stuck_run_steps": 2.6499998569488525, "final_distance": 4.301638603210449, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 7, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.308901+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.156155109405518, "chaser_new_fields": 33.97999954223633, "chaser_partner_vision": 0.0, "chaser_return": -3.2619988918304443, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 1.8600000143051147, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 74.72000122070312, "explorer_partner_vision": 0.0, "explorer_return": 61.44999694824219, "explorer_stuck_run_steps": 0.05999999865889549, "final_distance": 5.338278293609619, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 8, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.409994+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.038296699523926, "chaser_new_fields": 78.72000122070312, "chaser_partner_vision": 0.0, "chaser_return": 5.573996543884277, "chaser_stuck_run_steps": 0.04999999701976776, "collisions_per_episode": 2.7899999618530273, "degenerate_episodes": 85, "degenerate_fraction": 0.85, "episodes": 100, "explorer_new_fields": 32.939998626708984, "explorer_partner_vision": 0.0, "explorer_return": -1.25, "explorer_stuck_run_steps": 2.700000047683716, "final_distance": 4.791200637817383, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 9, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.506454+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.115623474121094, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.0, "chaser_return": -3.289998769760132, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 2.919999837875366, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 76.08000183105469, "explorer_partner_vision": 0.0, "explorer_return": 63.32499694824219, "explorer_stuck_run_steps": 0.07000000029802322, "final_distance": 5.588272571563721, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 10, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.606931+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.116372585296631, "chaser_new_fields": 76.8699951171875, "chaser_partner_vision": 0.0, "chaser_return": 5.221996784210205, "chaser_stuck_run_steps": 0.9899999499320984, "collisions_per_episode": 2.9099998474121094, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 32.80999755859375, "explorer_partner_vision": 0.0, "explorer_return": -1.5199999809265137, "explorer_stuck_run_steps": 2.7300000190734863, "final_distance": 4.800493240356445, "same_state_run_steps": 0.09999999403953552}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 11, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.703008+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.121584415435791, "chaser_new_fields": 33.97999954223633, "chaser_partner_vision": 0.0, "chaser_return": -3.2599987983703613, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 2.3499999046325684, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 71.93999481201172, "explorer_partner_vision": 0.0, "explorer_return": 57.11499786376953, "explorer_stuck_run_steps": 1.559999942779541, "final_distance": 4.933297634124756, "same_state_run_steps": 0.11999999731779099}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 12, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.802941+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.091162204742432, "chaser_new_fields": 77.73999786376953, "chaser_partner_vision": 0.0, "chaser_return": 5.407996654510498, "chaser_stuck_run_steps": 0.09999999403953552, "collisions_per_episode": 2.509999990463257, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 32.80999755859375, "explorer_partner_vision": 0.0, "explorer_return": -1.3399999141693115, "explorer_stuck_run_steps": 2.700000047683716, "final_distance": 5.153555870056152, "same_state_run_steps": 0.019999999552965164}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 13, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.900898+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.003658771514893, "chaser_new_fields": 33.75, "chaser_partner_vision": 0.0, "chaser_return": -3.3079988956451416, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 2.4800000190734863, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 60.68000030517578, "explorer_partner_vision": 0.0, "explorer_return": 40.404998779296875, "explorer_stuck_run_steps": 0.3199999928474426, "final_distance": 4.824435710906982, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 14, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.001633+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.128373146057129, "chaser_new_fields": 82.11000061035156, "chaser_partner_vision": 0.0, "chaser_return": 6.269996166229248, "chaser_stuck_run_steps": 0.08999999612569809, "collisions_per_episode": 2.93999981880188, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.040000915527344, "explorer_partner_vision": 0.0, "explorer_return": -0.9799999594688416, "explorer_stuck_run_steps": 2.7200000286102295, "final_distance": 5.028230667114258, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 15, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.097744+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.059652328491211, "chaser_new_fields": 33.790000915527344, "chaser_partner_vision": 0.0, "chaser_return": -3.2999987602233887, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 2.2899999618530273, "degenerate_episodes": 90, "degenerate_fraction": 0.9, "episodes": 100, "explorer_new_fields": 61.14999771118164, "explorer_partner_vision": 0.0, "explorer_return": 41.125, "explorer_stuck_run_steps": 0.19999998807907104, "final_distance": 4.463276386260986, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 16, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.197891+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.262848854064941, "chaser_new_fields": 78.5, "chaser_partner_vision": 0.0, "chaser_return": 5.547996520996094, "chaser_stuck_run_steps": 0.03999999910593033, "collisions_per_episode": 2.440000057220459, "degenerate_episodes": 83, "degenerate_fraction": 0.83, "episodes": 100, "explorer_new_fields": 32.88999938964844, "explorer_partner_vision": 0.0, "explorer_return": -1.3700000047683716, "explorer_stuck_run_steps": 2.740000009536743, "final_distance": 5.147022724151611, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 17, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.294102+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.165109157562256, "chaser_new_fields": 33.48999786376953, "chaser_partner_vision": 0.0, "chaser_return": -3.343998908996582, "chaser_stuck_run_steps": 2.5799999237060547, "collisions_per_episode": 2.2699999809265137, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 68.97000122070312, "explorer_partner_vision": 0.0, "explorer_return": 52.70499801635742, "explorer_stuck_run_steps": 0.8499999642372131, "final_distance": 5.33642578125, "same_state_run_steps": 0.12999999523162842}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 18, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.394659+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.134922504425049, "chaser_new_fields": 82.80999755859375, "chaser_partner_vision": 0.0, "chaser_return": 6.391995906829834, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 2.819999933242798, "degenerate_episodes": 83, "degenerate_fraction": 0.83, "episodes": 100, "explorer_new_fields": 32.959999084472656, "explorer_partner_vision": 0.0, "explorer_return": -1.2649999856948853, "explorer_stuck_run_steps": 2.679999828338623, "final_distance": 5.392402648925781, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 19, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.491290+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 4.98706579208374, "chaser_new_fields": 33.81999969482422, "chaser_partner_vision": 0.0, "chaser_return": -3.3119988441467285, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 2.490000009536743, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 64.93999481201172, "explorer_partner_vision": 0.0, "explorer_return": 46.67499923706055, "explorer_stuck_run_steps": 0.029999999329447746, "final_distance": 4.802732467651367, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 20, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.600352+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.9226698875427246, "chaser_new_fields": 31.739999771118164, "chaser_partner_vision": 0.9387999773025513, "chaser_return": 22.954994201660156, "chaser_stuck_run_steps": 0.6100000143051147, "collisions_per_episode": 24.729999542236328, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 27.779998779296875, "explorer_partner_vision": 0.9387999773025513, "explorer_return": -26.604999542236328, "explorer_stuck_run_steps": 3.259999990463257, "final_distance": 1.7273223400115967, "same_state_run_steps": 0.26999998092651367}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 21, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.704418+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.418818473815918, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.2913999855518341, "chaser_return": -1.88499915599823, "chaser_stuck_run_steps": 2.7100000381469727, "collisions_per_episode": 1.25, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 55.38999938964844, "explorer_partner_vision": 0.2913999855518341, "explorer_return": 32.06999969482422, "explorer_stuck_run_steps": 0.12999999523162842, "final_distance": 5.25391149520874, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 22, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.812676+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 2.0339229106903076, "chaser_new_fields": 33.209999084472656, "chaser_partner_vision": 0.9183999300003052, "chaser_return": 21.927995681762695, "chaser_stuck_run_steps": 0.6699999570846558, "collisions_per_episode": 23.51999855041504, "degenerate_episodes": 92, "degenerate_fraction": 0.92, "episodes": 100, "explorer_new_fields": 26.84000015258789, "explorer_partner_vision": 0.9183999300003052, "explorer_return": -26.510000228881836, "explorer_stuck_run_steps": 3.169999837875366, "final_distance": 1.7653462886810303, "same_state_run_steps": 0.25}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 23, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.916857+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.981630325317383, "chaser_new_fields": 33.93000030517578, "chaser_partner_vision": 0.2247999906539917, "chaser_return": -2.4529991149902344, "chaser_stuck_run_steps": 2.690000057220459, "collisions_per_episode": 0.7099999785423279, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 53.73999786376953, "explorer_partner_vision": 0.2247999906539917, "explorer_return": 29.98499870300293, "explorer_stuck_run_steps": 0.28999999165534973, "final_distance": 5.55134916305542, "same_state_run_steps": 0.03999999910593033}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 24, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.025324+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.846373200416565, "chaser_new_fields": 30.559999465942383, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 25.47699737548828, "chaser_stuck_run_steps": 2.4099998474121094, "collisions_per_episode": 27.189998626708984, "degenerate_episodes": 95, "degenerate_fraction": 0.95, "episodes": 100, "explorer_new_fields": 25.549999237060547, "explorer_partner_vision": 0.9455000162124634, "explorer_return": -29.889999389648438, "explorer_stuck_run_steps": 3.359999895095825, "final_distance": 1.641021728515625, "same_state_run_steps": 0.5399999618530273}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 25, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.131142+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.628232479095459, "chaser_new_fields": 33.95000076293945, "chaser_partner_vision": 0.26339998841285706, "chaser_return": -2.1799991130828857, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.9599999785423279, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 54.0, "explorer_partner_vision": 0.26339998841285706, "explorer_return": 30.189998626708984, "explorer_stuck_run_steps": 0.019999999552965164, "final_distance": 5.8266167640686035, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 26, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.255109+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.8304731845855713, "chaser_new_fields": 30.989999771118164, "chaser_partner_vision": 0.9462000131607056, "chaser_return": 27.23999786376953, "chaser_stuck_run_steps": 0.85999995470047, "collisions_per_episode": 28.760000228881836, "degenerate_episodes": 98, "degenerate_fraction": 0.98, "episodes": 100, "explorer_new_fields": 25.8799991607666, "explorer_partner_vision": 0.9462000131607056, "explorer_return": -31.03499984741211, "explorer_stuck_run_steps": 3.3499999046325684, "final_distance": 1.5262326002120972, "same_state_run_steps": 0.3799999952316284}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 27, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.376361+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.968653202056885, "chaser_new_fields": 33.86000061035156, "chaser_partner_vision": 0.22360000014305115, "chaser_return": -2.4379990100860596, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 0.7400000095367432, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 55.52000045776367, "explorer_partner_vision": 0.22360000014305115, "explorer_return": 32.59499740600586, "explorer_stuck_run_steps": 0.17999999225139618, "final_distance": 5.932736873626709, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 28, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.500759+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.7290464639663696, "chaser_new_fields": 28.920000076293945, "chaser_partner_vision": 0.9477999806404114, "chaser_return": 36.003997802734375, "chaser_stuck_run_steps": 1.3799999952316284, "collisions_per_episode": 37.13999938964844, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 23.78999900817871, "explorer_partner_vision": 0.9477999806404114, "explorer_return": -39.125, "explorer_stuck_run_steps": 3.5999999046325684, "final_distance": 1.566463589668274, "same_state_run_steps": 0.5299999713897705}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 29, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.610487+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.055323123931885, "chaser_new_fields": 33.90999984741211, "chaser_partner_vision": 0.204599991440773, "chaser_return": -2.4979991912841797, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.6800000071525574, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 46.849998474121094, "explorer_partner_vision": 0.204599991440773, "explorer_return": 19.69499969482422, "explorer_stuck_run_steps": 0.07000000029802322, "final_distance": 6.047131061553955, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 30, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.725728+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 2.022212505340576, "chaser_new_fields": 31.869998931884766, "chaser_partner_vision": 0.9148999452590942, "chaser_return": 23.19499397277832, "chaser_stuck_run_steps": 0.85999995470047, "collisions_per_episode": 24.889999389648438, "degenerate_episodes": 92, "degenerate_fraction": 0.92, "episodes": 100, "explorer_new_fields": 27.619998931884766, "explorer_partner_vision": 0.9148999452590942, "explorer_return": -25.889999389648438, "explorer_stuck_run_steps": 3.240000009536743, "final_distance": 1.8503373861312866, "same_state_run_steps": 0.25999999046325684}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 31, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.829596+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.011486530303955, "chaser_new_fields": 33.73999786376953, "chaser_partner_vision": 0.2134999781847, "chaser_return": -2.4539990425109863, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 0.7400000095367432, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 43.37999725341797, "explorer_partner_vision": 0.2134999781847, "explorer_return": 14.429999351501465, "explorer_stuck_run_steps": 0.17000000178813934, "final_distance": 5.857645034790039, "same_state_run_steps": 0.019999999552965164}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 32, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.937143+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.8730149269104004, "chaser_new_fields": 31.01999855041504, "chaser_partner_vision": 0.939300000667572, "chaser_return": 26.81099510192871, "chaser_stuck_run_steps": 0.7799999713897705, "collisions_per_episode": 28.369998931884766, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 25.760000228881836, "explorer_partner_vision": 0.939300000667572, "explorer_return": -30.899999618530273, "explorer_stuck_run_steps": 3.3299999237060547, "final_distance": 1.75227952003479, "same_state_run_steps": 0.2800000011920929}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 33, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.040804+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.9493937492370605, "chaser_new_fields": 33.93000030517578, "chaser_partner_vision": 0.21109998226165771, "chaser_return": -2.5089991092681885, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 0.6699999570846558, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 53.27000045776367, "explorer_partner_vision": 0.21109998226165771, "explorer_return": 29.344999313354492, "explorer_stuck_run_steps": 0.07000000029802322, "final_distance": 5.721303462982178, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 34, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.157423+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.8591976165771484, "chaser_new_fields": 30.69999885559082, "chaser_partner_vision": 0.9456999897956848, "chaser_return": 26.13799285888672, "chaser_stuck_run_steps": 1.0299999713897705, "collisions_per_episode": 27.799999237060547, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 26.239999771118164, "explorer_partner_vision": 0.9456999897956848, "explorer_return": -29.90999984741211, "explorer_stuck_run_steps": 3.369999885559082, "final_distance": 1.777151107788086, "same_state_run_steps": 0.4899999797344208}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 35, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.263149+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.806032657623291, "chaser_new_fields": 33.84000015258789, "chaser_partner_vision": 0.23879998922348022, "chaser_return": -2.4399991035461426, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 0.7400000095367432, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 57.55999755859375, "explorer_partner_vision": 0.23879998922348022, "explorer_return": 35.744998931884766, "explorer_stuck_run_steps": 0.08999999612569809, "final_distance": 6.035492897033691, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 36, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.371355+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.9622390270233154, "chaser_new_fields": 31.35999870300293, "chaser_partner_vision": 0.9269000291824341, "chaser_return": 23.311994552612305, "chaser_stuck_run_steps": 0.6299999952316284, "collisions_per_episode": 25.100000381469727, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 27.68000030517578, "explorer_partner_vision": 0.9269000291824341, "explorer_return": -26.309999465942383, "explorer_stuck_run_steps": 3.1399998664855957, "final_distance": 1.7391915321350098, "same_state_run_steps": 0.20999999344348907}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 37, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.475135+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.686277866363525, "chaser_new_fields": 33.880001068115234, "chaser_partner_vision": 0.2646999955177307, "chaser_return": -2.040999412536621, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 1.090000033378601, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 55.25, "explorer_partner_vision": 0.2646999955177307, "explorer_return": 31.864999771118164, "explorer_stuck_run_steps": 0.26999998092651367, "final_distance": 5.499191761016846, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 38, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.583483+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 2.13496994972229, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.8927000164985657, "chaser_return": 20.284996032714844, "chaser_stuck_run_steps": 0.550000011920929, "collisions_per_episode": 21.850000381469727, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 28.149999618530273, "explorer_partner_vision": 0.8927000164985657, "explorer_return": -23.260000228881836, "explorer_stuck_run_steps": 3.2200000286102295, "final_distance": 1.8080952167510986, "same_state_run_steps": 0.25}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 39, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.687176+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.822822093963623, "chaser_new_fields": 33.88999938964844, "chaser_partner_vision": 0.2704000174999237, "chaser_return": -1.4439992904663086, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 1.659999966621399, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 49.029998779296875, "explorer_partner_vision": 0.2704000174999237, "explorer_return": 22.23499870300293, "explorer_stuck_run_steps": 0.14999999105930328, "final_distance": 6.008201599121094, "same_state_run_steps": 0.029999999329447746}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 40, "source_path": "runs/mouse-run-run-2-mlp/paper_random_eval.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0009"} diff --git a/derived/tables/exclusions.jsonl b/derived/tables/exclusions.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/derived/tables/panels.json b/derived/tables/panels.json new file mode 100644 index 0000000000000000000000000000000000000000..72efb22e37cd89c6b88991937c0ebcc9a548b402 --- /dev/null +++ b/derived/tables/panels.json @@ -0,0 +1,28 @@ +{ + "panels": { + "paper_neural_behavior_non_degenerate_v1": { + "exclude_by": "exclusions.jsonl reason=paper_degenerate_episode_v1", + "table": "rollouts.jsonl", + "unit": "checkpoint rollout episode" + }, + "paper_random_opponent_valid_v1": { + "include": { + "evaluation_protocol": "paper_random_opponent_v1", + "status": "ok" + }, + "table": "evaluations.jsonl", + "unit": "checkpoint x opponent_mode" + }, + "primary_valid_runs_v1": { + "include": { + "checkpoint_ok": true, + "healthy": true, + "is_latest_successful": true, + "status": "completed" + }, + "table": "runs.jsonl", + "unit": "task x seed latest successful attempt" + } + }, + "schema_version": 1 +} diff --git a/derived/tables/rollouts.jsonl b/derived/tables/rollouts.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..c5dc5240b5af4a8c8a41197fa216d5111c76140d --- /dev/null +++ b/derived/tables/rollouts.jsonl @@ -0,0 +1,20 @@ +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:48.710464+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "source_line": 1, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:49.261015+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "source_line": 2, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:49.810154+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "source_line": 3, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:50.371542+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "source_line": 4, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:50.918340+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "source_line": 5, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:51.471824+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "source_line": 6, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:52.025242+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "source_line": 7, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:52.569627+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "source_line": 8, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:53.116716+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "source_line": 9, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:53.672368+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "source_line": 10, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:54.273478+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "source_line": 11, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:54.873436+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "source_line": 12, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:55.461961+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "source_line": 13, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:56.048411+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "source_line": 14, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:56.645581+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "source_line": 15, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:57.229171+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "source_line": 16, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:57.814519+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "source_line": 17, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:58.414469+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "source_line": 18, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:58.996792+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "source_line": 19, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:59.591392+00:00", "degenerate_episode_count": 0, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "source_line": 20, "source_path": "runs/mouse-run-run-2-mlp/paper_rollouts/paper_rollout_records.jsonl", "status": "ok", "task": "social", "unit_id": "social/seed_0009"} diff --git a/derived/tables/runs.jsonl b/derived/tables/runs.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..de87c2a41ba544214a890898f9504d84b57605cd --- /dev/null +++ b/derived/tables/runs.jsonl @@ -0,0 +1,20 @@ +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7602.132536457037, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "schema_version": 1, "seed": 0, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0000", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7618.575776674086, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "schema_version": 1, "seed": 1, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0001", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7622.409991315915, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "schema_version": 1, "seed": 2, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0002", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7631.6458840209525, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "schema_version": 1, "seed": 3, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0003", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7640.98172001401, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "schema_version": 1, "seed": 4, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0004", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7562.313855465036, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "schema_version": 1, "seed": 5, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0005", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7567.199341987027, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "schema_version": 1, "seed": 6, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0006", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7575.734206710942, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "schema_version": 1, "seed": 7, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0007", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7577.744137497037, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "schema_version": 1, "seed": 8, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0008", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7570.7650183150545, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "schema_version": 1, "seed": 9, "status": "completed", "task": "non_social", "total_updates": 20000, "unit_id": "non_social/seed_0009", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7633.100595657015, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "schema_version": 1, "seed": 0, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0000", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7644.564348316053, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "schema_version": 1, "seed": 1, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0001", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7640.485689831898, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "schema_version": 1, "seed": 2, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0002", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7642.0857080380665, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "schema_version": 1, "seed": 3, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0003", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7640.117992118001, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "schema_version": 1, "seed": 4, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0004", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7556.8758046310395, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "schema_version": 1, "seed": 5, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0005", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7558.8218611279735, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "schema_version": 1, "seed": 6, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0006", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7587.008876845008, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "schema_version": 1, "seed": 7, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0007", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7584.383489511092, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "schema_version": 1, "seed": 8, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0008", "update": 20000} +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_ok": true, "config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "elapsed_seconds": 7585.64040600392, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-mlp", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "schema_version": 1, "seed": 9, "status": "completed", "task": "social", "total_updates": 20000, "unit_id": "social/seed_0009", "update": 20000} diff --git a/eval.log b/eval.log new file mode 100644 index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 diff --git a/launch_gate.json b/launch_gate.json new file mode 100644 index 0000000000000000000000000000000000000000..f1d2ecdde86fe84b51dc397de8eeba619e0ce79b --- /dev/null +++ b/launch_gate.json @@ -0,0 +1,204 @@ +{ + "checks": [ + { + "detail": "experiments/paper_marl_2026/SPEC.md", + "name": "experiment_spec", + "ok": true + }, + { + "detail": { + "missing": [] + }, + "name": "config_schema", + "ok": true + }, + { + "detail": { + "actual": 20000, + "expected": 20000 + }, + "name": "primary_updates", + "ok": true + }, + { + "detail": { + "actual": 40, + "expected": 40 + }, + "name": "primary_batch_size", + "ok": true + }, + { + "detail": { + "actual": 100, + "expected": 100 + }, + "name": "primary_max_steps", + "ok": true + }, + { + "detail": { + "actual": 10, + "expected": 10 + }, + "name": "primary_seeds", + "ok": true + }, + { + "detail": { + "actual": 0, + "expected": 0 + }, + "name": "primary_seed_start", + "ok": true + }, + { + "detail": { + "actual": "both", + "expected": "both" + }, + "name": "primary_tasks", + "ok": true + }, + { + "detail": { + "actual": "paper_text", + "expected": "paper_text" + }, + "name": "primary_preset", + "ok": true + }, + { + "detail": { + "actual": "cuda", + "expected": "cuda" + }, + "name": "primary_device", + "ok": true + }, + { + "detail": { + "actual": true, + "expected": true + }, + "name": "primary_finite_guard", + "ok": true + }, + { + "detail": { + "actual": true, + "expected": true + }, + "name": "primary_triton_env_step", + "ok": true + }, + { + "detail": { + "actual": false, + "expected": false + }, + "name": "primary_fused_agent_rollout", + "ok": true + }, + { + "detail": { + "actual": 10, + "expected": 10 + }, + "name": "primary_subspace_metric_period", + "ok": true + }, + { + "detail": { + "actual": 2 + }, + "name": "max_attempts", + "ok": true + }, + { + "detail": { + "actual_seconds": 1800.0 + }, + "name": "observability_period", + "ok": true + }, + { + "detail": { + "actual_hours": 24.0 + }, + "name": "attempt_timeout", + "ok": true + }, + { + "detail": { + "cuda_available": true, + "device": "cuda" + }, + "name": "device_available", + "ok": true + }, + { + "detail": { + "current_dirty": false, + "current_git_sha": "6282c650d281b31d646d537dd3f57c968806a423", + "path": "runs/mouse-run-run-2-mlp/triton_equivalence.json", + "record_dirty": false, + "record_git_sha": "6282c650d281b31d646d537dd3f57c968806a423", + "sha_matches": true, + "status": "ok" + }, + "name": "triton_equivalence", + "ok": true + } + ], + "config": "experiments/paper_marl_2026/configs/arch_mlp.json", + "config_sha256": "4d12552c6b3cc4801827d0193b91d5ab945da76650f3691d3086729e2cc0c723", + "created_at": "2026-07-02T07:36:15.459261+00:00", + "experiment_spec": "experiments/paper_marl_2026/SPEC.md", + "experiment_spec_sha256": "e878bd03a06735c93c8caab4fc769ce63f026a538dfccbc87b588d432e48761e", + "provenance": { + "command": "scripts/run/launch_gate.py --config experiments/paper_marl_2026/configs/arch_mlp.json --triton-equivalence runs/mouse-run-run-2-mlp/triton_equivalence.json", + "created_at": "2026-07-02T07:36:15.459353+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "schema_version": 1, + "status": "ok", + "triton_equivalence": "runs/mouse-run-run-2-mlp/triton_equivalence.json" +} diff --git a/logs/non_social_seed_0000_attempt_01.log b/logs/non_social_seed_0000_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..4421cd3ec2e332e3b5682cfd300c799b04c4656c --- /dev/null +++ b/logs/non_social_seed_0000_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.521858+00:00 task=non_social seed=0 attempt=1 === +update=0001 collisions=1.70 chaser_return=-2.91 explorer_return=0.10 vision=(0.000,0.000) new_fields=(35.8,33.7) distance=5.47 loss=(-0.000,3.074) kl=-0.0000 +update=0500 collisions=3.38 chaser_return=4.70 explorer_return=49.94 vision=(0.000,0.000) new_fields=(74.2,67.3) distance=4.96 loss=(-0.002,4.558) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=1.90 chaser_return=5.81 explorer_return=50.42 vision=(0.000,0.000) new_fields=(79.5,67.3) distance=5.55 loss=(-0.002,4.399) kl=0.0007 +update=1500 collisions=3.05 chaser_return=6.08 explorer_return=53.65 vision=(0.000,0.000) new_fields=(80.8,69.6) distance=5.11 loss=(-0.001,3.947) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=2.12 chaser_return=5.82 explorer_return=50.42 vision=(0.000,0.000) new_fields=(79.7,67.3) distance=4.40 loss=(-0.001,4.047) kl=0.0008 +update=2500 collisions=2.30 chaser_return=5.18 explorer_return=55.11 vision=(0.000,0.000) new_fields=(76.7,70.8) distance=5.46 loss=(-0.001,3.958) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=2.38 chaser_return=5.36 explorer_return=50.58 vision=(0.000,0.000) new_fields=(77.6,67.6) distance=5.60 loss=(-0.002,3.992) kl=0.0004 +update=3500 collisions=1.90 chaser_return=6.02 explorer_return=54.36 vision=(0.000,0.000) new_fields=(80.7,70.0) distance=4.94 loss=(-0.000,3.771) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=1.65 chaser_return=5.88 explorer_return=57.17 vision=(0.000,0.000) new_fields=(79.8,71.9) distance=5.85 loss=(-0.001,3.668) kl=0.0001 +update=4500 collisions=1.62 chaser_return=6.29 explorer_return=54.14 vision=(0.000,0.000) new_fields=(81.8,69.8) distance=5.53 loss=(-0.001,3.734) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=2.15 chaser_return=5.38 explorer_return=59.09 vision=(0.000,0.000) new_fields=(77.4,73.2) distance=5.58 loss=(-0.000,3.802) kl=0.0008 +update=5500 collisions=1.62 chaser_return=5.69 explorer_return=58.53 vision=(0.000,0.000) new_fields=(78.8,72.8) distance=5.20 loss=(-0.001,3.841) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.85 chaser_return=5.56 explorer_return=57.70 vision=(0.000,0.000) new_fields=(78.3,72.4) distance=4.69 loss=(-0.001,3.698) kl=0.0004 +update=6500 collisions=1.73 chaser_return=6.44 explorer_return=58.60 vision=(0.000,0.000) new_fields=(82.6,72.9) distance=5.05 loss=(-0.001,3.696) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=1.60 chaser_return=6.43 explorer_return=65.50 vision=(0.000,0.000) new_fields=(82.6,77.5) distance=4.87 loss=(-0.000,3.672) kl=-0.0000 +update=7500 collisions=1.77 chaser_return=5.95 explorer_return=65.05 vision=(0.000,0.000) new_fields=(80.2,77.2) distance=5.50 loss=(-0.001,3.700) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=1.65 chaser_return=5.95 explorer_return=63.25 vision=(0.000,0.000) new_fields=(80.3,76.0) distance=5.81 loss=(-0.000,3.573) kl=0.0000 +update=8500 collisions=1.93 chaser_return=6.19 explorer_return=64.56 vision=(0.000,0.000) new_fields=(81.4,76.7) distance=5.31 loss=(-0.001,3.583) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=1.83 chaser_return=5.74 explorer_return=66.33 vision=(0.000,0.000) new_fields=(79.3,78.0) distance=4.99 loss=(-0.001,3.665) kl=0.0005 +update=9500 collisions=1.83 chaser_return=6.16 explorer_return=64.94 vision=(0.000,0.000) new_fields=(81.3,77.1) distance=4.54 loss=(-0.000,3.484) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=2.25 chaser_return=6.16 explorer_return=65.28 vision=(0.000,0.000) new_fields=(81.4,77.2) distance=5.32 loss=(-0.000,3.523) kl=-0.0002 +update=10500 collisions=2.00 chaser_return=6.42 explorer_return=65.20 vision=(0.000,0.000) new_fields=(82.7,77.1) distance=5.14 loss=(-0.001,3.464) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=1.52 chaser_return=6.52 explorer_return=66.59 vision=(0.000,0.000) new_fields=(83.1,78.2) distance=4.93 loss=(-0.000,3.494) kl=0.0001 +update=11500 collisions=1.40 chaser_return=6.07 explorer_return=69.33 vision=(0.000,0.000) new_fields=(80.8,80.0) distance=5.81 loss=(-0.002,3.456) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=1.65 chaser_return=5.11 explorer_return=66.18 vision=(0.000,0.000) new_fields=(75.9,78.0) distance=5.17 loss=(-0.001,3.763) kl=0.0008 +update=12500 collisions=1.35 chaser_return=6.79 explorer_return=68.12 vision=(0.000,0.000) new_fields=(84.3,79.1) distance=5.73 loss=(-0.001,3.352) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=1.40 chaser_return=6.55 explorer_return=67.64 vision=(0.000,0.000) new_fields=(83.2,78.7) distance=5.54 loss=(-0.001,3.472) kl=0.0005 +update=13500 collisions=1.73 chaser_return=6.42 explorer_return=69.70 vision=(0.000,0.000) new_fields=(82.6,80.3) distance=5.57 loss=(-0.002,3.334) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=1.83 chaser_return=6.42 explorer_return=66.62 vision=(0.000,0.000) new_fields=(82.8,78.2) distance=5.71 loss=(-0.001,3.345) kl=0.0002 +update=14500 collisions=1.70 chaser_return=6.39 explorer_return=68.01 vision=(0.000,0.000) new_fields=(82.6,79.1) distance=5.15 loss=(-0.000,3.397) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=1.40 chaser_return=6.77 explorer_return=67.53 vision=(0.000,0.000) new_fields=(84.5,78.7) distance=5.50 loss=(-0.000,3.233) kl=0.0003 +update=15500 collisions=1.18 chaser_return=6.56 explorer_return=63.85 vision=(0.000,0.000) new_fields=(83.1,76.2) distance=5.34 loss=(-0.001,3.442) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=1.70 chaser_return=6.42 explorer_return=69.96 vision=(0.000,0.000) new_fields=(82.7,80.5) distance=5.65 loss=(-0.001,3.293) kl=0.0001 +update=16500 collisions=1.58 chaser_return=6.63 explorer_return=71.50 vision=(0.000,0.000) new_fields=(83.6,81.5) distance=5.41 loss=(-0.000,3.280) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=1.48 chaser_return=6.36 explorer_return=70.04 vision=(0.000,0.000) new_fields=(82.3,80.5) distance=5.11 loss=(-0.000,3.309) kl=0.0002 +update=17500 collisions=1.35 chaser_return=6.46 explorer_return=71.80 vision=(0.000,0.000) new_fields=(82.8,81.5) distance=5.24 loss=(-0.000,3.166) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=1.18 chaser_return=6.41 explorer_return=68.84 vision=(0.000,0.000) new_fields=(82.4,79.7) distance=5.66 loss=(-0.000,3.226) kl=0.0005 +update=18500 collisions=1.52 chaser_return=6.22 explorer_return=69.55 vision=(0.000,0.000) new_fields=(81.5,80.0) distance=6.22 loss=(-0.001,3.249) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=1.30 chaser_return=6.22 explorer_return=65.99 vision=(0.000,0.000) new_fields=(81.6,77.7) distance=5.97 loss=(-0.000,3.290) kl=-0.0000 +update=19500 collisions=1.52 chaser_return=6.90 explorer_return=64.11 vision=(0.000,0.000) new_fields=(85.1,76.4) distance=6.13 loss=(-0.001,3.472) kl=0.0001 +update=20000 collisions=1.25 chaser_return=6.54 explorer_return=62.76 vision=(0.000,0.000) new_fields=(83.1,75.5) distance=5.31 loss=(-0.001,3.419) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7602.184716115007, "metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "task": "non_social", "unit_id": "non_social/seed_0000"} +=== exited 2026-07-02T09:43:08.049146+00:00 returncode=0 === diff --git a/logs/non_social_seed_0001_attempt_01.log b/logs/non_social_seed_0001_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..241cf589a2848f9ea1356f217183ee3338c35ac4 --- /dev/null +++ b/logs/non_social_seed_0001_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.523172+00:00 task=non_social seed=1 attempt=1 === +update=0001 collisions=1.10 chaser_return=-3.14 explorer_return=4.04 vision=(0.000,0.000) new_fields=(34.5,36.2) distance=5.17 loss=(-0.001,2.900) kl=-0.0000 +update=0500 collisions=1.05 chaser_return=5.33 explorer_return=48.62 vision=(0.000,0.000) new_fields=(76.9,66.1) distance=5.66 loss=(-0.002,4.656) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=1.08 chaser_return=5.80 explorer_return=50.28 vision=(0.000,0.000) new_fields=(79.2,67.0) distance=5.04 loss=(-0.002,4.247) kl=0.0003 +update=1500 collisions=1.25 chaser_return=6.27 explorer_return=47.73 vision=(0.000,0.000) new_fields=(81.6,65.5) distance=4.56 loss=(-0.001,4.188) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=1.30 chaser_return=6.26 explorer_return=50.95 vision=(0.000,0.000) new_fields=(81.6,67.7) distance=5.20 loss=(-0.000,3.860) kl=0.0006 +update=2500 collisions=1.30 chaser_return=6.16 explorer_return=52.45 vision=(0.000,0.000) new_fields=(81.2,68.6) distance=4.25 loss=(-0.001,4.090) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=1.80 chaser_return=6.49 explorer_return=56.39 vision=(0.000,0.000) new_fields=(82.9,71.2) distance=4.91 loss=(-0.000,3.652) kl=0.0005 +update=3500 collisions=1.38 chaser_return=6.53 explorer_return=56.05 vision=(0.000,0.000) new_fields=(83.1,71.0) distance=4.81 loss=(-0.001,3.655) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=1.73 chaser_return=6.30 explorer_return=57.29 vision=(0.000,0.000) new_fields=(81.9,71.9) distance=4.88 loss=(-0.001,3.659) kl=-0.0001 +update=4500 collisions=1.48 chaser_return=6.14 explorer_return=52.26 vision=(0.000,0.000) new_fields=(81.1,68.6) distance=4.94 loss=(-0.000,3.886) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=1.18 chaser_return=5.84 explorer_return=53.01 vision=(0.000,0.000) new_fields=(79.7,69.0) distance=5.93 loss=(-0.000,3.810) kl=0.0002 +update=5500 collisions=1.68 chaser_return=5.90 explorer_return=56.95 vision=(0.000,0.000) new_fields=(79.9,71.6) distance=4.47 loss=(-0.001,3.734) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.25 chaser_return=5.76 explorer_return=55.41 vision=(0.000,0.000) new_fields=(79.1,70.6) distance=5.37 loss=(-0.000,3.706) kl=0.0001 +update=6500 collisions=1.40 chaser_return=5.33 explorer_return=60.40 vision=(0.000,0.000) new_fields=(76.9,74.0) distance=5.48 loss=(-0.000,3.598) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=1.18 chaser_return=5.19 explorer_return=62.61 vision=(0.000,0.000) new_fields=(76.4,75.4) distance=6.20 loss=(-0.001,4.010) kl=0.0001 +update=7500 collisions=1.02 chaser_return=6.66 explorer_return=63.06 vision=(0.000,0.000) new_fields=(83.7,75.6) distance=5.51 loss=(-0.000,3.763) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=1.05 chaser_return=6.45 explorer_return=63.70 vision=(0.000,0.000) new_fields=(82.4,76.2) distance=5.56 loss=(-0.001,3.566) kl=0.0005 +update=8500 collisions=1.00 chaser_return=6.19 explorer_return=65.16 vision=(0.000,0.000) new_fields=(81.2,77.2) distance=5.27 loss=(-0.001,3.467) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=0.85 chaser_return=6.68 explorer_return=65.76 vision=(0.000,0.000) new_fields=(83.7,77.4) distance=5.15 loss=(-0.000,3.594) kl=0.0001 +update=9500 collisions=1.00 chaser_return=4.38 explorer_return=59.46 vision=(0.000,0.000) new_fields=(72.2,73.3) distance=5.29 loss=(-0.000,3.870) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=1.02 chaser_return=6.09 explorer_return=62.31 vision=(0.000,0.000) new_fields=(80.7,75.3) distance=5.23 loss=(-0.001,3.616) kl=0.0004 +update=10500 collisions=1.43 chaser_return=5.69 explorer_return=58.98 vision=(0.000,0.000) new_fields=(78.9,73.2) distance=4.75 loss=(-0.001,3.461) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=2.23 chaser_return=6.07 explorer_return=58.98 vision=(0.000,0.000) new_fields=(80.7,73.1) distance=4.46 loss=(-0.001,3.696) kl=0.0002 +update=11500 collisions=1.40 chaser_return=5.59 explorer_return=65.58 vision=(0.000,0.000) new_fields=(78.5,77.5) distance=5.03 loss=(-0.001,3.508) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=1.60 chaser_return=5.33 explorer_return=54.36 vision=(0.000,0.000) new_fields=(77.0,70.1) distance=5.15 loss=(-0.001,3.632) kl=0.0002 +update=12500 collisions=1.90 chaser_return=5.13 explorer_return=55.60 vision=(0.000,0.000) new_fields=(76.1,70.9) distance=4.85 loss=(-0.001,3.693) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=1.15 chaser_return=5.46 explorer_return=61.19 vision=(0.000,0.000) new_fields=(77.6,74.6) distance=5.08 loss=(-0.001,3.686) kl=0.0008 +update=13500 collisions=1.70 chaser_return=5.43 explorer_return=56.28 vision=(0.000,0.000) new_fields=(77.5,71.2) distance=5.20 loss=(-0.001,3.767) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=2.03 chaser_return=5.36 explorer_return=56.58 vision=(0.000,0.000) new_fields=(77.4,71.5) distance=4.74 loss=(-0.001,3.665) kl=0.0002 +update=14500 collisions=1.93 chaser_return=5.19 explorer_return=51.85 vision=(0.000,0.000) new_fields=(76.5,68.4) distance=5.30 loss=(-0.002,3.699) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=1.77 chaser_return=5.49 explorer_return=58.60 vision=(0.000,0.000) new_fields=(77.9,72.9) distance=5.04 loss=(-0.001,3.715) kl=0.0002 +update=15500 collisions=1.35 chaser_return=5.79 explorer_return=60.29 vision=(0.000,0.000) new_fields=(79.3,73.9) distance=5.41 loss=(-0.001,3.430) kl=0.0015 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=1.35 chaser_return=5.40 explorer_return=57.51 vision=(0.000,0.000) new_fields=(77.3,72.0) distance=5.17 loss=(-0.000,3.629) kl=0.0003 +update=16500 collisions=1.93 chaser_return=5.49 explorer_return=53.28 vision=(0.000,0.000) new_fields=(77.9,69.4) distance=5.08 loss=(-0.001,3.668) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=1.88 chaser_return=5.45 explorer_return=58.00 vision=(0.000,0.000) new_fields=(77.9,72.5) distance=5.44 loss=(-0.001,3.733) kl=0.0002 +update=17500 collisions=1.83 chaser_return=5.95 explorer_return=60.33 vision=(0.000,0.000) new_fields=(80.2,74.2) distance=5.85 loss=(-0.000,3.431) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=1.58 chaser_return=5.58 explorer_return=62.24 vision=(0.000,0.000) new_fields=(78.2,75.2) distance=5.50 loss=(-0.001,3.572) kl=0.0002 +update=18500 collisions=1.48 chaser_return=5.70 explorer_return=51.70 vision=(0.000,0.000) new_fields=(79.0,68.2) distance=4.39 loss=(-0.001,3.885) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=1.50 chaser_return=6.04 explorer_return=58.30 vision=(0.000,0.000) new_fields=(80.8,72.6) distance=4.76 loss=(-0.000,3.455) kl=0.0002 +update=19500 collisions=0.90 chaser_return=6.27 explorer_return=57.10 vision=(0.000,0.000) new_fields=(81.6,71.7) distance=4.83 loss=(-0.000,3.564) kl=0.0006 +update=20000 collisions=0.98 chaser_return=6.60 explorer_return=59.76 vision=(0.000,0.000) new_fields=(83.2,73.5) distance=4.44 loss=(-0.001,3.403) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7618.63883137505, "metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "task": "non_social", "unit_id": "non_social/seed_0001"} +=== exited 2026-07-02T09:43:18.367158+00:00 returncode=0 === diff --git a/logs/non_social_seed_0002_attempt_01.log b/logs/non_social_seed_0002_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..7942b8be77baa9aed246572e07b9f301e22af6e2 --- /dev/null +++ b/logs/non_social_seed_0002_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.524044+00:00 task=non_social seed=2 attempt=1 === +update=0001 collisions=1.80 chaser_return=-3.30 explorer_return=4.97 vision=(0.000,0.000) new_fields=(33.8,36.8) distance=5.51 loss=(-0.001,3.103) kl=0.0000 +update=0500 collisions=2.25 chaser_return=4.99 explorer_return=44.84 vision=(0.000,0.000) new_fields=(75.3,63.8) distance=5.45 loss=(-0.001,4.764) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=2.62 chaser_return=5.26 explorer_return=47.69 vision=(0.000,0.000) new_fields=(76.8,65.6) distance=5.27 loss=(-0.002,4.376) kl=0.0007 +update=1500 collisions=2.58 chaser_return=4.76 explorer_return=51.78 vision=(0.000,0.000) new_fields=(74.4,68.5) distance=5.16 loss=(-0.002,4.042) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=3.25 chaser_return=4.16 explorer_return=53.50 vision=(0.000,0.000) new_fields=(71.5,69.5) distance=5.11 loss=(-0.001,4.174) kl=0.0000 +update=2500 collisions=2.45 chaser_return=4.94 explorer_return=60.44 vision=(0.000,0.000) new_fields=(75.4,74.1) distance=5.53 loss=(-0.001,3.834) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=2.20 chaser_return=5.45 explorer_return=61.75 vision=(0.000,0.000) new_fields=(77.8,75.0) distance=5.20 loss=(-0.001,3.790) kl=-0.0001 +update=3500 collisions=2.40 chaser_return=5.00 explorer_return=55.98 vision=(0.000,0.000) new_fields=(75.6,71.1) distance=5.27 loss=(-0.002,4.077) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=3.20 chaser_return=5.04 explorer_return=56.99 vision=(0.000,0.000) new_fields=(76.1,71.9) distance=5.15 loss=(-0.001,3.932) kl=0.0001 +update=4500 collisions=2.03 chaser_return=4.83 explorer_return=61.15 vision=(0.000,0.000) new_fields=(74.7,74.6) distance=4.90 loss=(-0.001,3.991) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=3.03 chaser_return=5.47 explorer_return=57.48 vision=(0.000,0.000) new_fields=(78.2,72.2) distance=4.81 loss=(-0.001,3.771) kl=0.0002 +update=5500 collisions=2.08 chaser_return=5.41 explorer_return=60.59 vision=(0.000,0.000) new_fields=(77.6,74.3) distance=4.74 loss=(-0.001,3.858) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=2.38 chaser_return=5.44 explorer_return=59.88 vision=(0.000,0.000) new_fields=(77.9,73.6) distance=4.70 loss=(-0.000,4.051) kl=0.0000 +update=6500 collisions=2.33 chaser_return=5.87 explorer_return=68.01 vision=(0.000,0.000) new_fields=(79.9,79.1) distance=5.67 loss=(-0.000,3.768) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=2.62 chaser_return=4.46 explorer_return=69.36 vision=(0.000,0.000) new_fields=(73.0,80.2) distance=5.27 loss=(-0.001,3.815) kl=0.0006 +update=7500 collisions=2.75 chaser_return=5.92 explorer_return=66.74 vision=(0.000,0.000) new_fields=(80.3,78.4) distance=4.29 loss=(-0.001,3.648) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=2.35 chaser_return=5.92 explorer_return=69.29 vision=(0.000,0.000) new_fields=(80.1,79.9) distance=4.07 loss=(-0.001,3.647) kl=0.0003 +update=8500 collisions=2.33 chaser_return=5.90 explorer_return=70.30 vision=(0.000,0.000) new_fields=(80.1,80.6) distance=3.76 loss=(-0.001,3.613) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=1.93 chaser_return=6.26 explorer_return=75.10 vision=(0.000,0.000) new_fields=(81.8,83.8) distance=4.52 loss=(-0.000,3.443) kl=0.0006 +update=9500 collisions=2.40 chaser_return=6.10 explorer_return=74.88 vision=(0.000,0.000) new_fields=(81.0,83.7) distance=4.31 loss=(-0.001,3.433) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=2.12 chaser_return=5.89 explorer_return=73.41 vision=(0.000,0.000) new_fields=(80.0,82.7) distance=4.59 loss=(-0.000,3.523) kl=-0.0001 +update=10500 collisions=2.78 chaser_return=5.69 explorer_return=61.38 vision=(0.000,0.000) new_fields=(79.1,74.8) distance=5.01 loss=(-0.001,3.884) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=2.23 chaser_return=6.36 explorer_return=61.90 vision=(0.000,0.000) new_fields=(82.2,75.0) distance=4.50 loss=(-0.001,3.602) kl=0.0003 +update=11500 collisions=1.43 chaser_return=6.49 explorer_return=71.76 vision=(0.000,0.000) new_fields=(82.9,81.5) distance=4.82 loss=(-0.000,3.341) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=2.80 chaser_return=6.28 explorer_return=64.75 vision=(0.000,0.000) new_fields=(82.1,77.0) distance=4.86 loss=(-0.002,3.599) kl=0.0006 +update=12500 collisions=1.93 chaser_return=6.36 explorer_return=66.93 vision=(0.000,0.000) new_fields=(82.2,78.4) distance=4.70 loss=(-0.000,3.529) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=1.75 chaser_return=6.35 explorer_return=69.21 vision=(0.000,0.000) new_fields=(82.2,79.9) distance=4.77 loss=(-0.001,3.496) kl=0.0005 +update=13500 collisions=1.98 chaser_return=6.57 explorer_return=66.70 vision=(0.000,0.000) new_fields=(83.6,78.1) distance=5.22 loss=(-0.000,3.530) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=2.58 chaser_return=6.49 explorer_return=71.84 vision=(0.000,0.000) new_fields=(83.1,81.9) distance=5.08 loss=(-0.000,3.381) kl=-0.0000 +update=14500 collisions=1.90 chaser_return=6.67 explorer_return=70.08 vision=(0.000,0.000) new_fields=(83.9,80.5) distance=4.94 loss=(-0.000,3.423) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=2.08 chaser_return=6.53 explorer_return=69.03 vision=(0.000,0.000) new_fields=(83.1,79.8) distance=4.93 loss=(-0.001,3.598) kl=-0.0000 +update=15500 collisions=2.20 chaser_return=6.24 explorer_return=70.00 vision=(0.000,0.000) new_fields=(81.9,80.7) distance=4.71 loss=(-0.001,3.464) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=1.50 chaser_return=6.61 explorer_return=70.94 vision=(0.000,0.000) new_fields=(83.4,81.0) distance=5.13 loss=(-0.000,3.484) kl=0.0000 +update=16500 collisions=2.00 chaser_return=6.35 explorer_return=70.15 vision=(0.000,0.000) new_fields=(82.5,80.5) distance=4.24 loss=(-0.000,3.485) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=2.10 chaser_return=6.20 explorer_return=69.21 vision=(0.000,0.000) new_fields=(81.5,79.8) distance=4.89 loss=(-0.001,3.600) kl=0.0006 +update=17500 collisions=1.93 chaser_return=6.54 explorer_return=66.93 vision=(0.000,0.000) new_fields=(83.1,78.3) distance=5.16 loss=(-0.001,3.538) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=1.98 chaser_return=5.89 explorer_return=69.14 vision=(0.000,0.000) new_fields=(80.0,79.8) distance=5.03 loss=(-0.001,3.525) kl=0.0009 +update=18500 collisions=4.85 chaser_return=5.94 explorer_return=66.55 vision=(0.000,0.000) new_fields=(80.2,78.0) distance=4.18 loss=(-0.001,3.637) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=2.40 chaser_return=6.31 explorer_return=69.29 vision=(0.000,0.000) new_fields=(82.1,80.1) distance=4.77 loss=(-0.000,3.486) kl=0.0004 +update=19500 collisions=2.42 chaser_return=6.49 explorer_return=72.51 vision=(0.000,0.000) new_fields=(83.0,82.1) distance=5.00 loss=(-0.000,3.395) kl=0.0001 +update=20000 collisions=2.08 chaser_return=6.76 explorer_return=73.30 vision=(0.000,0.000) new_fields=(84.3,82.8) distance=4.67 loss=(-0.001,3.510) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7622.430775379995, "metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "task": "non_social", "unit_id": "non_social/seed_0002"} +=== exited 2026-07-02T09:43:28.528507+00:00 returncode=0 === diff --git a/logs/non_social_seed_0003_attempt_01.log b/logs/non_social_seed_0003_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..36abe930aa128b2328259674f51620790303950c --- /dev/null +++ b/logs/non_social_seed_0003_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.524887+00:00 task=non_social seed=3 attempt=1 === +update=0001 collisions=2.10 chaser_return=-3.79 explorer_return=-0.57 vision=(0.000,0.000) new_fields=(31.4,33.2) distance=4.79 loss=(-0.000,3.267) kl=0.0000 +update=0500 collisions=3.30 chaser_return=5.14 explorer_return=48.55 vision=(0.000,0.000) new_fields=(76.2,66.1) distance=6.01 loss=(-0.002,4.514) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=4.10 chaser_return=5.51 explorer_return=46.67 vision=(0.000,0.000) new_fields=(78.3,65.2) distance=4.67 loss=(-0.002,4.498) kl=0.0004 +update=1500 collisions=2.48 chaser_return=6.05 explorer_return=43.34 vision=(0.000,0.000) new_fields=(80.9,62.7) distance=5.01 loss=(-0.001,4.305) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=2.33 chaser_return=5.94 explorer_return=42.21 vision=(0.000,0.000) new_fields=(80.6,61.9) distance=5.07 loss=(-0.001,4.197) kl=0.0005 +update=2500 collisions=2.38 chaser_return=6.28 explorer_return=43.23 vision=(0.000,0.000) new_fields=(82.1,62.6) distance=5.09 loss=(-0.000,4.095) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=2.33 chaser_return=6.37 explorer_return=43.08 vision=(0.000,0.000) new_fields=(82.6,62.5) distance=5.17 loss=(-0.001,3.895) kl=0.0002 +update=3500 collisions=2.50 chaser_return=6.58 explorer_return=46.08 vision=(0.000,0.000) new_fields=(83.7,64.7) distance=6.12 loss=(-0.001,3.854) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=2.28 chaser_return=6.61 explorer_return=45.48 vision=(0.000,0.000) new_fields=(83.8,64.0) distance=5.39 loss=(-0.000,3.766) kl=-0.0002 +update=4500 collisions=1.93 chaser_return=6.38 explorer_return=44.73 vision=(0.000,0.000) new_fields=(82.7,63.5) distance=5.78 loss=(-0.001,3.812) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=1.73 chaser_return=6.71 explorer_return=43.98 vision=(0.000,0.000) new_fields=(84.2,63.0) distance=6.08 loss=(-0.000,3.856) kl=0.0001 +update=5500 collisions=2.35 chaser_return=6.35 explorer_return=47.01 vision=(0.000,0.000) new_fields=(82.4,65.1) distance=6.20 loss=(-0.001,3.659) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=2.25 chaser_return=6.50 explorer_return=49.38 vision=(0.000,0.000) new_fields=(83.2,66.8) distance=5.63 loss=(-0.001,3.732) kl=0.0002 +update=6500 collisions=1.83 chaser_return=6.26 explorer_return=45.66 vision=(0.000,0.000) new_fields=(81.9,64.3) distance=6.03 loss=(-0.000,3.905) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=2.28 chaser_return=6.52 explorer_return=49.94 vision=(0.000,0.000) new_fields=(83.4,67.1) distance=5.44 loss=(-0.001,3.666) kl=0.0006 +update=7500 collisions=1.73 chaser_return=6.92 explorer_return=58.75 vision=(0.000,0.000) new_fields=(85.2,72.9) distance=5.51 loss=(-0.001,3.484) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=2.67 chaser_return=6.61 explorer_return=56.50 vision=(0.000,0.000) new_fields=(83.9,71.6) distance=5.52 loss=(-0.000,3.613) kl=0.0003 +update=8500 collisions=1.50 chaser_return=6.84 explorer_return=55.45 vision=(0.000,0.000) new_fields=(84.6,70.6) distance=5.27 loss=(-0.001,3.571) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=1.60 chaser_return=6.97 explorer_return=57.17 vision=(0.000,0.000) new_fields=(85.3,71.8) distance=6.14 loss=(-0.001,3.552) kl=0.0001 +update=9500 collisions=1.62 chaser_return=7.01 explorer_return=58.23 vision=(0.000,0.000) new_fields=(85.6,72.5) distance=5.88 loss=(-0.001,3.518) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=1.65 chaser_return=6.77 explorer_return=54.14 vision=(0.000,0.000) new_fields=(84.4,69.8) distance=5.77 loss=(-0.001,3.711) kl=0.0004 +update=10500 collisions=2.00 chaser_return=6.75 explorer_return=54.81 vision=(0.000,0.000) new_fields=(84.3,70.3) distance=5.62 loss=(-0.001,3.519) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=1.50 chaser_return=6.85 explorer_return=57.17 vision=(0.000,0.000) new_fields=(84.8,71.9) distance=5.25 loss=(0.000,3.642) kl=-0.0000 +update=11500 collisions=1.38 chaser_return=7.05 explorer_return=54.33 vision=(0.000,0.000) new_fields=(85.6,70.0) distance=5.77 loss=(-0.001,3.390) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=1.77 chaser_return=7.19 explorer_return=54.14 vision=(0.000,0.000) new_fields=(86.6,69.8) distance=6.10 loss=(-0.001,3.461) kl=0.0003 +update=12500 collisions=1.50 chaser_return=7.08 explorer_return=53.39 vision=(0.000,0.000) new_fields=(85.8,69.4) distance=5.96 loss=(-0.001,3.413) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=1.68 chaser_return=7.19 explorer_return=53.20 vision=(0.000,0.000) new_fields=(86.6,69.3) distance=6.22 loss=(-0.001,3.548) kl=0.0001 +update=13500 collisions=2.00 chaser_return=7.20 explorer_return=50.84 vision=(0.000,0.000) new_fields=(86.7,67.7) distance=5.75 loss=(-0.000,3.510) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=3.50 chaser_return=6.98 explorer_return=50.20 vision=(0.000,0.000) new_fields=(85.6,67.4) distance=4.92 loss=(-0.001,3.714) kl=0.0000 +update=14500 collisions=2.12 chaser_return=6.22 explorer_return=46.08 vision=(0.000,0.000) new_fields=(81.8,64.5) distance=3.76 loss=(-0.001,3.703) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=1.88 chaser_return=6.36 explorer_return=48.62 vision=(0.000,0.000) new_fields=(82.3,66.1) distance=4.35 loss=(-0.001,3.744) kl=0.0000 +update=15500 collisions=1.98 chaser_return=6.39 explorer_return=40.26 vision=(0.000,0.000) new_fields=(82.6,60.7) distance=4.36 loss=(-0.001,3.929) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=1.90 chaser_return=6.93 explorer_return=49.94 vision=(0.000,0.000) new_fields=(85.2,67.0) distance=4.74 loss=(-0.001,3.416) kl=0.0002 +update=16500 collisions=1.77 chaser_return=6.88 explorer_return=49.90 vision=(0.000,0.000) new_fields=(84.9,67.0) distance=4.16 loss=(-0.001,3.580) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=1.62 chaser_return=7.02 explorer_return=61.79 vision=(0.000,0.000) new_fields=(85.6,75.1) distance=4.57 loss=(-0.001,3.388) kl=0.0002 +update=17500 collisions=2.05 chaser_return=7.17 explorer_return=51.29 vision=(0.000,0.000) new_fields=(86.5,68.1) distance=4.23 loss=(-0.001,3.639) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=2.67 chaser_return=6.88 explorer_return=54.51 vision=(0.000,0.000) new_fields=(85.2,70.3) distance=4.54 loss=(-0.001,3.795) kl=0.0003 +update=18500 collisions=2.00 chaser_return=7.42 explorer_return=58.86 vision=(0.000,0.000) new_fields=(87.7,73.1) distance=3.79 loss=(-0.001,3.571) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=2.15 chaser_return=7.32 explorer_return=60.21 vision=(0.000,0.000) new_fields=(87.3,74.0) distance=4.64 loss=(-0.000,3.456) kl=0.0005 +update=19500 collisions=1.90 chaser_return=7.38 explorer_return=62.76 vision=(0.000,0.000) new_fields=(87.6,75.7) distance=3.78 loss=(-0.001,3.499) kl=0.0006 +update=20000 collisions=1.73 chaser_return=7.30 explorer_return=62.12 vision=(0.000,0.000) new_fields=(87.1,75.2) distance=4.14 loss=(-0.001,3.491) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7631.695038600941, "metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "task": "non_social", "unit_id": "non_social/seed_0003"} +=== exited 2026-07-02T09:43:38.602587+00:00 returncode=0 === diff --git a/logs/non_social_seed_0004_attempt_01.log b/logs/non_social_seed_0004_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..e257887e54fd196004c66c2ceb2e11a90111f3fd --- /dev/null +++ b/logs/non_social_seed_0004_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.525938+00:00 task=non_social seed=4 attempt=1 === +update=0001 collisions=1.52 chaser_return=-2.70 explorer_return=2.12 vision=(0.000,0.000) new_fields=(36.8,35.0) distance=5.16 loss=(-0.001,3.118) kl=-0.0000 +update=0500 collisions=2.80 chaser_return=4.54 explorer_return=44.31 vision=(0.000,0.000) new_fields=(73.3,63.2) distance=5.77 loss=(-0.002,4.773) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=3.12 chaser_return=4.64 explorer_return=46.60 vision=(0.000,0.000) new_fields=(73.9,65.0) distance=5.48 loss=(-0.001,4.439) kl=0.0003 +update=1500 collisions=2.67 chaser_return=5.74 explorer_return=51.51 vision=(0.000,0.000) new_fields=(79.4,68.2) distance=5.40 loss=(-0.001,4.171) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=2.62 chaser_return=5.74 explorer_return=47.46 vision=(0.000,0.000) new_fields=(79.3,65.6) distance=4.74 loss=(-0.001,4.247) kl=0.0001 +update=2500 collisions=2.03 chaser_return=5.46 explorer_return=46.90 vision=(0.000,0.000) new_fields=(77.7,65.3) distance=5.53 loss=(-0.001,4.049) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=3.23 chaser_return=5.51 explorer_return=47.58 vision=(0.000,0.000) new_fields=(78.2,65.7) distance=5.26 loss=(-0.002,4.019) kl=0.0002 +update=3500 collisions=1.75 chaser_return=6.24 explorer_return=51.29 vision=(0.000,0.000) new_fields=(81.8,68.0) distance=5.07 loss=(-0.001,3.856) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=1.80 chaser_return=5.94 explorer_return=53.16 vision=(0.000,0.000) new_fields=(80.2,69.2) distance=4.02 loss=(-0.000,3.730) kl=0.0000 +update=4500 collisions=1.88 chaser_return=6.07 explorer_return=54.40 vision=(0.000,0.000) new_fields=(80.8,70.1) distance=5.64 loss=(-0.001,3.669) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=2.05 chaser_return=6.00 explorer_return=56.16 vision=(0.000,0.000) new_fields=(80.7,71.1) distance=4.47 loss=(-0.001,3.657) kl=0.0005 +update=5500 collisions=2.33 chaser_return=6.25 explorer_return=57.17 vision=(0.000,0.000) new_fields=(81.8,72.0) distance=4.90 loss=(-0.001,3.611) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.80 chaser_return=6.38 explorer_return=57.36 vision=(0.000,0.000) new_fields=(82.5,72.0) distance=4.56 loss=(-0.001,3.647) kl=0.0004 +update=6500 collisions=2.48 chaser_return=5.81 explorer_return=57.70 vision=(0.000,0.000) new_fields=(79.7,72.3) distance=5.06 loss=(-0.001,3.639) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=1.93 chaser_return=6.27 explorer_return=57.74 vision=(0.000,0.000) new_fields=(82.0,72.4) distance=4.56 loss=(-0.000,3.609) kl=-0.0001 +update=7500 collisions=2.00 chaser_return=5.67 explorer_return=56.01 vision=(0.000,0.000) new_fields=(78.9,71.2) distance=5.03 loss=(-0.001,3.669) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=2.30 chaser_return=6.02 explorer_return=55.38 vision=(0.000,0.000) new_fields=(80.9,70.7) distance=5.81 loss=(-0.001,3.602) kl=0.0003 +update=8500 collisions=4.60 chaser_return=4.99 explorer_return=50.16 vision=(0.000,0.000) new_fields=(75.7,67.3) distance=4.70 loss=(-0.001,3.843) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=2.70 chaser_return=5.79 explorer_return=54.81 vision=(0.000,0.000) new_fields=(79.7,70.5) distance=4.68 loss=(-0.000,3.554) kl=0.0002 +update=9500 collisions=2.12 chaser_return=5.53 explorer_return=53.35 vision=(0.000,0.000) new_fields=(78.2,69.5) distance=5.71 loss=(-0.001,3.767) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=2.30 chaser_return=5.39 explorer_return=58.38 vision=(0.000,0.000) new_fields=(77.6,72.9) distance=5.03 loss=(-0.001,3.573) kl=0.0008 +update=10500 collisions=1.27 chaser_return=5.81 explorer_return=58.53 vision=(0.000,0.000) new_fields=(79.4,72.8) distance=5.08 loss=(-0.001,3.605) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=2.23 chaser_return=5.72 explorer_return=59.54 vision=(0.000,0.000) new_fields=(79.3,73.5) distance=5.25 loss=(-0.001,3.539) kl=-0.0002 +update=11500 collisions=1.88 chaser_return=5.56 explorer_return=57.03 vision=(0.000,0.000) new_fields=(78.4,71.9) distance=5.20 loss=(-0.001,3.630) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=2.08 chaser_return=5.09 explorer_return=55.75 vision=(0.000,0.000) new_fields=(75.9,71.1) distance=5.03 loss=(-0.001,3.760) kl=0.0006 +update=12500 collisions=1.95 chaser_return=6.06 explorer_return=56.46 vision=(0.000,0.000) new_fields=(80.9,71.3) distance=5.68 loss=(-0.001,3.521) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=2.10 chaser_return=5.33 explorer_return=54.92 vision=(0.000,0.000) new_fields=(77.2,70.5) distance=5.22 loss=(-0.001,3.856) kl=0.0009 +update=13500 collisions=2.23 chaser_return=5.47 explorer_return=59.76 vision=(0.000,0.000) new_fields=(78.0,73.6) distance=4.86 loss=(-0.001,3.574) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=1.90 chaser_return=5.82 explorer_return=60.44 vision=(0.000,0.000) new_fields=(79.5,74.0) distance=4.30 loss=(-0.001,3.546) kl=0.0004 +update=14500 collisions=2.28 chaser_return=5.39 explorer_return=59.42 vision=(0.000,0.000) new_fields=(77.4,73.5) distance=5.56 loss=(-0.000,3.604) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=1.85 chaser_return=6.00 explorer_return=58.19 vision=(0.000,0.000) new_fields=(80.5,72.6) distance=4.86 loss=(-0.001,3.492) kl=0.0001 +update=15500 collisions=1.93 chaser_return=6.20 explorer_return=57.81 vision=(0.000,0.000) new_fields=(81.8,72.4) distance=5.56 loss=(-0.001,3.532) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=2.00 chaser_return=6.21 explorer_return=57.51 vision=(0.000,0.000) new_fields=(81.6,72.1) distance=4.76 loss=(-0.001,3.554) kl=0.0004 +update=16500 collisions=2.03 chaser_return=5.92 explorer_return=59.84 vision=(0.000,0.000) new_fields=(80.4,73.8) distance=4.73 loss=(-0.001,3.486) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=1.60 chaser_return=5.77 explorer_return=57.81 vision=(0.000,0.000) new_fields=(79.4,72.3) distance=5.04 loss=(-0.001,3.596) kl=0.0001 +update=17500 collisions=1.55 chaser_return=6.05 explorer_return=61.90 vision=(0.000,0.000) new_fields=(80.6,75.1) distance=5.21 loss=(-0.001,3.360) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=2.35 chaser_return=5.90 explorer_return=63.17 vision=(0.000,0.000) new_fields=(80.1,76.1) distance=5.27 loss=(-0.001,3.491) kl=0.0002 +update=18500 collisions=1.60 chaser_return=6.21 explorer_return=63.40 vision=(0.000,0.000) new_fields=(81.5,76.1) distance=5.32 loss=(-0.001,3.353) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=1.68 chaser_return=5.94 explorer_return=61.30 vision=(0.000,0.000) new_fields=(80.2,74.7) distance=5.57 loss=(-0.001,3.479) kl=0.0003 +update=19500 collisions=2.23 chaser_return=5.84 explorer_return=62.58 vision=(0.000,0.000) new_fields=(79.8,75.6) distance=5.48 loss=(-0.000,3.403) kl=0.0004 +update=20000 collisions=1.83 chaser_return=6.33 explorer_return=66.10 vision=(0.000,0.000) new_fields=(82.2,78.0) distance=5.31 loss=(-0.001,3.323) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7640.99740531703, "metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "task": "non_social", "unit_id": "non_social/seed_0004"} +=== exited 2026-07-02T09:43:48.667830+00:00 returncode=0 === diff --git a/logs/non_social_seed_0005_attempt_01.log b/logs/non_social_seed_0005_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..f4c4ccf42f2b74ef6b0bbe9a452b11a3380258c7 --- /dev/null +++ b/logs/non_social_seed_0005_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:28.527565+00:00 task=non_social seed=5 attempt=1 === +update=0001 collisions=1.38 chaser_return=-3.43 explorer_return=1.86 vision=(0.000,0.000) new_fields=(33.1,34.8) distance=5.59 loss=(-0.001,3.093) kl=0.0000 +update=0500 collisions=1.48 chaser_return=5.37 explorer_return=54.59 vision=(0.000,0.000) new_fields=(77.4,70.1) distance=4.62 loss=(-0.002,4.404) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=1.60 chaser_return=5.86 explorer_return=56.39 vision=(0.000,0.000) new_fields=(79.8,71.3) distance=5.76 loss=(-0.001,4.144) kl=0.0001 +update=1500 collisions=0.73 chaser_return=6.31 explorer_return=56.24 vision=(0.000,0.000) new_fields=(81.8,71.0) distance=5.41 loss=(-0.001,3.992) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=1.35 chaser_return=5.83 explorer_return=56.01 vision=(0.000,0.000) new_fields=(79.5,71.0) distance=4.97 loss=(-0.001,4.067) kl=0.0005 +update=2500 collisions=0.82 chaser_return=6.57 explorer_return=60.06 vision=(0.000,0.000) new_fields=(83.0,73.6) distance=4.89 loss=(-0.002,4.075) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=1.18 chaser_return=6.21 explorer_return=59.69 vision=(0.000,0.000) new_fields=(81.2,73.5) distance=4.73 loss=(-0.001,4.040) kl=-0.0001 +update=3500 collisions=1.33 chaser_return=6.47 explorer_return=61.23 vision=(0.000,0.000) new_fields=(82.8,74.6) distance=4.77 loss=(-0.001,3.945) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=1.75 chaser_return=6.95 explorer_return=61.23 vision=(0.000,0.000) new_fields=(85.3,74.7) distance=4.62 loss=(-0.000,3.940) kl=-0.0001 +update=4500 collisions=1.43 chaser_return=6.96 explorer_return=60.89 vision=(0.000,0.000) new_fields=(85.2,74.2) distance=5.47 loss=(-0.001,3.911) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=1.20 chaser_return=6.60 explorer_return=61.34 vision=(0.000,0.000) new_fields=(83.3,74.5) distance=4.56 loss=(-0.001,3.877) kl=0.0002 +update=5500 collisions=0.93 chaser_return=5.47 explorer_return=61.79 vision=(0.000,0.000) new_fields=(77.6,74.8) distance=5.23 loss=(-0.001,3.934) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.02 chaser_return=5.31 explorer_return=60.21 vision=(0.000,0.000) new_fields=(77.0,73.7) distance=5.47 loss=(-0.000,3.929) kl=0.0001 +update=6500 collisions=1.18 chaser_return=4.91 explorer_return=59.05 vision=(0.000,0.000) new_fields=(75.0,73.0) distance=4.71 loss=(-0.001,3.863) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=1.38 chaser_return=5.28 explorer_return=63.17 vision=(0.000,0.000) new_fields=(76.8,75.8) distance=5.45 loss=(-0.000,3.864) kl=0.0001 +update=7500 collisions=1.43 chaser_return=5.53 explorer_return=64.90 vision=(0.000,0.000) new_fields=(78.0,77.0) distance=4.91 loss=(-0.001,3.865) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=1.68 chaser_return=5.52 explorer_return=61.45 vision=(0.000,0.000) new_fields=(78.0,74.7) distance=4.85 loss=(-0.001,3.723) kl=0.0004 +update=8500 collisions=1.08 chaser_return=5.36 explorer_return=65.43 vision=(0.000,0.000) new_fields=(77.0,77.3) distance=4.80 loss=(-0.001,3.755) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=1.10 chaser_return=5.75 explorer_return=63.70 vision=(0.000,0.000) new_fields=(78.9,76.2) distance=5.00 loss=(-0.001,3.859) kl=0.0004 +update=9500 collisions=0.98 chaser_return=6.59 explorer_return=61.19 vision=(0.000,0.000) new_fields=(83.2,74.5) distance=4.30 loss=(-0.001,3.872) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=1.05 chaser_return=6.06 explorer_return=64.00 vision=(0.000,0.000) new_fields=(80.5,76.3) distance=5.21 loss=(-0.001,3.785) kl=0.0003 +update=10500 collisions=0.90 chaser_return=5.39 explorer_return=63.17 vision=(0.000,0.000) new_fields=(77.2,75.7) distance=5.27 loss=(-0.000,3.767) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=1.18 chaser_return=5.79 explorer_return=61.23 vision=(0.000,0.000) new_fields=(79.2,74.4) distance=5.14 loss=(-0.000,3.715) kl=0.0001 +update=11500 collisions=1.30 chaser_return=6.08 explorer_return=61.23 vision=(0.000,0.000) new_fields=(80.8,74.6) distance=5.31 loss=(-0.001,3.808) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=1.20 chaser_return=5.33 explorer_return=64.83 vision=(0.000,0.000) new_fields=(77.0,76.8) distance=4.53 loss=(-0.001,3.822) kl=0.0001 +update=12500 collisions=1.33 chaser_return=5.14 explorer_return=63.44 vision=(0.000,0.000) new_fields=(76.0,76.0) distance=5.12 loss=(-0.000,3.905) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=0.82 chaser_return=5.30 explorer_return=61.34 vision=(0.000,0.000) new_fields=(76.8,74.5) distance=4.83 loss=(-0.001,3.705) kl=-0.0002 +update=13500 collisions=0.70 chaser_return=4.78 explorer_return=64.71 vision=(0.000,0.000) new_fields=(74.2,76.7) distance=4.84 loss=(-0.001,3.741) kl=-0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=0.85 chaser_return=4.31 explorer_return=60.48 vision=(0.000,0.000) new_fields=(71.8,74.0) distance=4.77 loss=(-0.001,3.648) kl=0.0004 +update=14500 collisions=0.95 chaser_return=4.47 explorer_return=61.53 vision=(0.000,0.000) new_fields=(72.7,74.6) distance=4.75 loss=(-0.000,3.682) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=1.43 chaser_return=4.67 explorer_return=62.54 vision=(0.000,0.000) new_fields=(73.7,75.4) distance=4.49 loss=(-0.001,3.678) kl=0.0005 +update=15500 collisions=1.25 chaser_return=4.96 explorer_return=59.58 vision=(0.000,0.000) new_fields=(75.2,73.5) distance=4.57 loss=(-0.001,3.831) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=1.10 chaser_return=4.91 explorer_return=61.56 vision=(0.000,0.000) new_fields=(74.8,74.8) distance=3.97 loss=(-0.000,3.616) kl=0.0004 +update=16500 collisions=1.43 chaser_return=4.88 explorer_return=59.31 vision=(0.000,0.000) new_fields=(74.9,73.4) distance=4.76 loss=(-0.000,3.797) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=1.62 chaser_return=5.41 explorer_return=58.67 vision=(0.000,0.000) new_fields=(77.6,73.0) distance=5.04 loss=(-0.001,3.581) kl=0.0003 +update=17500 collisions=1.12 chaser_return=4.96 explorer_return=62.99 vision=(0.000,0.000) new_fields=(75.1,75.8) distance=5.48 loss=(-0.001,3.771) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=1.23 chaser_return=5.24 explorer_return=61.41 vision=(0.000,0.000) new_fields=(76.6,74.7) distance=4.87 loss=(-0.001,3.698) kl=0.0004 +update=18500 collisions=1.52 chaser_return=5.74 explorer_return=61.08 vision=(0.000,0.000) new_fields=(79.2,74.5) distance=5.09 loss=(-0.001,3.678) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=1.33 chaser_return=5.32 explorer_return=62.61 vision=(0.000,0.000) new_fields=(76.9,75.5) distance=4.53 loss=(-0.000,3.703) kl=0.0003 +update=19500 collisions=0.73 chaser_return=5.46 explorer_return=63.33 vision=(0.000,0.000) new_fields=(77.5,75.7) distance=4.44 loss=(-0.001,3.721) kl=0.0006 +update=20000 collisions=1.02 chaser_return=6.02 explorer_return=59.42 vision=(0.000,0.000) new_fields=(80.5,73.2) distance=4.73 loss=(-0.000,3.523) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7562.32880273601, "metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "task": "non_social", "unit_id": "non_social/seed_0005"} +=== exited 2026-07-02T11:49:40.894281+00:00 returncode=0 === diff --git a/logs/non_social_seed_0006_attempt_01.log b/logs/non_social_seed_0006_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..44d44a8efc3f177baa357ca8d6aeca5c06c17250 --- /dev/null +++ b/logs/non_social_seed_0006_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:48.617227+00:00 task=non_social seed=6 attempt=1 === +update=0001 collisions=2.35 chaser_return=-3.20 explorer_return=-0.43 vision=(0.000,0.000) new_fields=(34.4,33.5) distance=4.61 loss=(-0.001,3.030) kl=0.0000 +update=0500 collisions=0.75 chaser_return=5.09 explorer_return=43.26 vision=(0.000,0.000) new_fields=(75.6,62.5) distance=5.97 loss=(-0.002,4.700) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=0.95 chaser_return=5.83 explorer_return=40.90 vision=(0.000,0.000) new_fields=(79.4,60.9) distance=5.96 loss=(-0.001,4.234) kl=0.0002 +update=1500 collisions=0.90 chaser_return=6.05 explorer_return=38.84 vision=(0.000,0.000) new_fields=(80.5,59.5) distance=5.93 loss=(-0.001,4.259) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=0.60 chaser_return=5.84 explorer_return=39.36 vision=(0.000,0.000) new_fields=(79.3,59.7) distance=5.52 loss=(-0.001,4.038) kl=0.0002 +update=2500 collisions=0.53 chaser_return=6.87 explorer_return=41.58 vision=(0.000,0.000) new_fields=(84.5,61.2) distance=5.54 loss=(-0.000,3.781) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=0.62 chaser_return=6.37 explorer_return=43.83 vision=(0.000,0.000) new_fields=(82.1,62.7) distance=5.62 loss=(-0.001,4.088) kl=-0.0001 +update=3500 collisions=0.65 chaser_return=6.87 explorer_return=38.95 vision=(0.000,0.000) new_fields=(84.6,59.4) distance=5.97 loss=(-0.000,3.927) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=0.88 chaser_return=6.80 explorer_return=43.75 vision=(0.000,0.000) new_fields=(84.2,62.7) distance=5.37 loss=(-0.001,3.941) kl=-0.0000 +update=4500 collisions=0.62 chaser_return=6.63 explorer_return=41.61 vision=(0.000,0.000) new_fields=(83.4,61.2) distance=5.82 loss=(-0.001,3.979) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=1.43 chaser_return=6.47 explorer_return=38.35 vision=(0.000,0.000) new_fields=(82.8,59.4) distance=4.96 loss=(-0.001,3.918) kl=0.0001 +update=5500 collisions=1.10 chaser_return=6.12 explorer_return=42.81 vision=(0.000,0.000) new_fields=(81.0,62.1) distance=4.76 loss=(-0.001,3.662) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.23 chaser_return=6.65 explorer_return=41.76 vision=(0.000,0.000) new_fields=(83.6,61.6) distance=5.43 loss=(-0.001,3.932) kl=0.0006 +update=6500 collisions=0.78 chaser_return=6.62 explorer_return=45.03 vision=(0.000,0.000) new_fields=(83.5,63.6) distance=5.08 loss=(-0.000,3.829) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=1.27 chaser_return=6.36 explorer_return=45.10 vision=(0.000,0.000) new_fields=(82.3,63.7) distance=5.44 loss=(-0.001,3.815) kl=0.0003 +update=7500 collisions=1.05 chaser_return=5.68 explorer_return=46.53 vision=(0.000,0.000) new_fields=(78.7,64.6) distance=5.28 loss=(-0.001,3.874) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=0.70 chaser_return=5.67 explorer_return=46.19 vision=(0.000,0.000) new_fields=(78.5,64.4) distance=4.73 loss=(-0.001,3.969) kl=0.0007 +update=8500 collisions=0.98 chaser_return=6.50 explorer_return=53.05 vision=(0.000,0.000) new_fields=(83.0,68.9) distance=4.71 loss=(-0.001,3.682) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=1.10 chaser_return=5.61 explorer_return=55.11 vision=(0.000,0.000) new_fields=(78.4,70.5) distance=4.92 loss=(-0.001,3.816) kl=0.0003 +update=9500 collisions=0.70 chaser_return=6.09 explorer_return=48.96 vision=(0.000,0.000) new_fields=(80.6,66.2) distance=5.17 loss=(-0.001,3.700) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=0.70 chaser_return=5.81 explorer_return=54.25 vision=(0.000,0.000) new_fields=(79.3,69.6) distance=4.84 loss=(0.000,3.720) kl=0.0002 +update=10500 collisions=1.08 chaser_return=6.25 explorer_return=52.86 vision=(0.000,0.000) new_fields=(81.6,69.0) distance=4.62 loss=(-0.001,3.482) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=0.55 chaser_return=6.15 explorer_return=49.90 vision=(0.000,0.000) new_fields=(80.8,66.7) distance=4.71 loss=(-0.001,3.714) kl=0.0001 +update=11500 collisions=1.12 chaser_return=6.43 explorer_return=38.09 vision=(0.000,0.000) new_fields=(82.5,58.9) distance=4.79 loss=(-0.001,4.131) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=0.82 chaser_return=6.47 explorer_return=37.45 vision=(0.000,0.000) new_fields=(82.6,58.4) distance=4.44 loss=(-0.001,3.982) kl=0.0005 +update=12500 collisions=0.85 chaser_return=6.77 explorer_return=42.03 vision=(0.000,0.000) new_fields=(84.2,61.5) distance=4.87 loss=(-0.001,3.788) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=1.05 chaser_return=5.94 explorer_return=44.54 vision=(0.000,0.000) new_fields=(80.1,63.2) distance=5.31 loss=(-0.001,4.082) kl=0.0004 +update=13500 collisions=1.35 chaser_return=6.58 explorer_return=44.24 vision=(0.000,0.000) new_fields=(83.4,63.2) distance=4.13 loss=(-0.000,3.821) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=1.98 chaser_return=6.18 explorer_return=40.94 vision=(0.000,0.000) new_fields=(81.5,61.1) distance=4.75 loss=(-0.001,4.174) kl=0.0006 +update=14500 collisions=1.55 chaser_return=6.05 explorer_return=40.45 vision=(0.000,0.000) new_fields=(80.6,60.6) distance=4.70 loss=(-0.000,3.985) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=1.15 chaser_return=6.25 explorer_return=38.69 vision=(0.000,0.000) new_fields=(81.5,59.5) distance=4.01 loss=(-0.000,3.674) kl=0.0004 +update=15500 collisions=1.27 chaser_return=6.11 explorer_return=41.31 vision=(0.000,0.000) new_fields=(80.9,61.2) distance=3.81 loss=(-0.001,3.731) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=1.93 chaser_return=5.63 explorer_return=43.41 vision=(0.000,0.000) new_fields=(78.7,62.7) distance=4.86 loss=(-0.001,3.888) kl=0.0003 +update=16500 collisions=1.20 chaser_return=6.00 explorer_return=40.64 vision=(0.000,0.000) new_fields=(80.2,60.7) distance=4.30 loss=(-0.000,3.691) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=1.12 chaser_return=5.92 explorer_return=40.75 vision=(0.000,0.000) new_fields=(79.9,60.9) distance=4.73 loss=(-0.001,3.881) kl=0.0006 +update=17500 collisions=0.95 chaser_return=6.14 explorer_return=41.35 vision=(0.000,0.000) new_fields=(81.0,61.3) distance=4.91 loss=(-0.000,3.892) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=0.93 chaser_return=6.10 explorer_return=42.92 vision=(0.000,0.000) new_fields=(80.8,62.2) distance=4.83 loss=(-0.001,3.887) kl=0.0004 +update=18500 collisions=1.25 chaser_return=6.06 explorer_return=42.96 vision=(0.000,0.000) new_fields=(80.7,62.3) distance=4.71 loss=(-0.000,3.709) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=1.58 chaser_return=6.40 explorer_return=41.65 vision=(0.000,0.000) new_fields=(82.4,61.5) distance=4.66 loss=(-0.001,3.625) kl=0.0004 +update=19500 collisions=1.65 chaser_return=6.14 explorer_return=42.66 vision=(0.000,0.000) new_fields=(81.4,62.2) distance=4.72 loss=(-0.001,3.940) kl=0.0006 +update=20000 collisions=1.15 chaser_return=6.39 explorer_return=41.24 vision=(0.000,0.000) new_fields=(82.4,61.0) distance=4.67 loss=(-0.000,3.841) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7567.213006881066, "metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "task": "non_social", "unit_id": "non_social/seed_0006"} +=== exited 2026-07-02T11:50:00.915352+00:00 returncode=0 === diff --git a/logs/non_social_seed_0007_attempt_01.log b/logs/non_social_seed_0007_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..98917532a34cfe80866fc50495dd26979a95da73 --- /dev/null +++ b/logs/non_social_seed_0007_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:58.681826+00:00 task=non_social seed=7 attempt=1 === +update=0001 collisions=1.48 chaser_return=-3.24 explorer_return=1.49 vision=(0.000,0.000) new_fields=(34.1,34.5) distance=5.13 loss=(-0.000,3.129) kl=0.0000 +update=0500 collisions=3.03 chaser_return=4.67 explorer_return=45.03 vision=(0.000,0.000) new_fields=(74.1,64.0) distance=5.01 loss=(-0.002,4.623) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=2.88 chaser_return=4.99 explorer_return=46.23 vision=(0.000,0.000) new_fields=(75.7,64.7) distance=5.35 loss=(-0.001,4.559) kl=0.0003 +update=1500 collisions=2.53 chaser_return=5.85 explorer_return=51.25 vision=(0.000,0.000) new_fields=(79.9,68.1) distance=4.94 loss=(-0.001,4.427) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=2.88 chaser_return=5.68 explorer_return=56.42 vision=(0.000,0.000) new_fields=(79.4,71.6) distance=6.11 loss=(-0.001,4.207) kl=0.0001 +update=2500 collisions=1.88 chaser_return=5.87 explorer_return=61.38 vision=(0.000,0.000) new_fields=(79.8,74.8) distance=4.83 loss=(-0.001,3.954) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=2.40 chaser_return=4.89 explorer_return=62.54 vision=(0.000,0.000) new_fields=(75.1,75.5) distance=4.51 loss=(-0.000,4.105) kl=0.0001 +update=3500 collisions=2.42 chaser_return=6.16 explorer_return=59.39 vision=(0.000,0.000) new_fields=(81.4,73.4) distance=5.17 loss=(-0.001,3.993) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=2.55 chaser_return=5.89 explorer_return=65.65 vision=(0.000,0.000) new_fields=(80.1,77.8) distance=5.10 loss=(-0.001,3.974) kl=0.0002 +update=4500 collisions=2.55 chaser_return=5.20 explorer_return=61.83 vision=(0.000,0.000) new_fields=(76.5,75.1) distance=5.03 loss=(-0.000,4.066) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=2.28 chaser_return=4.79 explorer_return=64.53 vision=(0.000,0.000) new_fields=(74.7,76.8) distance=5.29 loss=(-0.001,3.989) kl=0.0005 +update=5500 collisions=2.70 chaser_return=5.46 explorer_return=68.16 vision=(0.000,0.000) new_fields=(78.1,79.5) distance=4.66 loss=(-0.001,3.927) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=2.25 chaser_return=5.89 explorer_return=65.91 vision=(0.000,0.000) new_fields=(80.2,77.7) distance=4.68 loss=(-0.001,3.832) kl=0.0004 +update=6500 collisions=2.33 chaser_return=5.18 explorer_return=63.92 vision=(0.000,0.000) new_fields=(76.4,76.3) distance=4.94 loss=(-0.001,4.087) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=2.17 chaser_return=5.18 explorer_return=66.21 vision=(0.000,0.000) new_fields=(76.4,78.0) distance=4.88 loss=(-0.001,4.070) kl=0.0006 +update=7500 collisions=4.03 chaser_return=5.55 explorer_return=62.01 vision=(0.000,0.000) new_fields=(78.5,75.2) distance=5.08 loss=(-0.001,3.999) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=2.62 chaser_return=5.84 explorer_return=68.09 vision=(0.000,0.000) new_fields=(80.1,79.1) distance=4.90 loss=(-0.001,3.641) kl=0.0002 +update=8500 collisions=2.60 chaser_return=6.58 explorer_return=63.62 vision=(0.000,0.000) new_fields=(83.7,76.2) distance=5.30 loss=(-0.001,3.789) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=2.05 chaser_return=5.99 explorer_return=60.66 vision=(0.000,0.000) new_fields=(80.6,74.3) distance=4.89 loss=(-0.001,3.983) kl=0.0002 +update=9500 collisions=2.48 chaser_return=6.74 explorer_return=62.46 vision=(0.000,0.000) new_fields=(84.5,75.3) distance=4.45 loss=(-0.001,3.623) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=3.20 chaser_return=5.66 explorer_return=61.19 vision=(0.000,0.000) new_fields=(79.1,74.7) distance=4.18 loss=(-0.001,3.959) kl=0.0001 +update=10500 collisions=2.75 chaser_return=6.05 explorer_return=61.23 vision=(0.000,0.000) new_fields=(81.0,74.7) distance=4.20 loss=(-0.000,3.825) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=2.48 chaser_return=6.60 explorer_return=58.56 vision=(0.000,0.000) new_fields=(83.6,72.8) distance=4.32 loss=(-0.000,3.884) kl=0.0002 +update=11500 collisions=2.62 chaser_return=6.16 explorer_return=66.81 vision=(0.000,0.000) new_fields=(81.5,78.3) distance=4.88 loss=(-0.000,3.705) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=2.62 chaser_return=6.47 explorer_return=67.75 vision=(0.000,0.000) new_fields=(83.0,79.0) distance=5.42 loss=(-0.001,3.711) kl=0.0005 +update=12500 collisions=2.23 chaser_return=5.84 explorer_return=61.90 vision=(0.000,0.000) new_fields=(79.7,75.0) distance=5.01 loss=(-0.001,3.898) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=2.73 chaser_return=5.98 explorer_return=63.74 vision=(0.000,0.000) new_fields=(80.5,76.3) distance=5.38 loss=(-0.001,3.841) kl=0.0003 +update=13500 collisions=2.05 chaser_return=6.51 explorer_return=71.61 vision=(0.000,0.000) new_fields=(83.3,81.6) distance=4.74 loss=(-0.000,3.638) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=2.67 chaser_return=6.09 explorer_return=74.01 vision=(0.000,0.000) new_fields=(81.1,83.3) distance=4.75 loss=(-0.000,3.677) kl=0.0001 +update=14500 collisions=3.23 chaser_return=5.49 explorer_return=67.41 vision=(0.000,0.000) new_fields=(78.4,79.1) distance=4.86 loss=(-0.001,3.808) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=2.67 chaser_return=5.91 explorer_return=73.30 vision=(0.000,0.000) new_fields=(80.3,83.0) distance=5.65 loss=(-0.001,3.792) kl=0.0002 +update=15500 collisions=2.12 chaser_return=5.64 explorer_return=70.15 vision=(0.000,0.000) new_fields=(78.7,80.6) distance=5.56 loss=(-0.000,3.819) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=2.00 chaser_return=5.39 explorer_return=67.53 vision=(0.000,0.000) new_fields=(77.5,79.0) distance=4.95 loss=(-0.000,3.818) kl=0.0003 +update=16500 collisions=2.60 chaser_return=5.80 explorer_return=69.74 vision=(0.000,0.000) new_fields=(79.6,80.5) distance=5.16 loss=(-0.000,3.713) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=3.85 chaser_return=6.39 explorer_return=62.16 vision=(0.000,0.000) new_fields=(82.7,75.3) distance=5.26 loss=(-0.001,3.738) kl=0.0000 +update=17500 collisions=2.75 chaser_return=5.80 explorer_return=67.45 vision=(0.000,0.000) new_fields=(79.5,78.7) distance=4.74 loss=(-0.001,3.659) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=2.65 chaser_return=6.38 explorer_return=64.97 vision=(0.000,0.000) new_fields=(82.7,77.3) distance=5.41 loss=(-0.000,3.824) kl=0.0001 +update=18500 collisions=2.75 chaser_return=6.22 explorer_return=63.59 vision=(0.000,0.000) new_fields=(81.9,76.1) distance=5.82 loss=(-0.001,3.752) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=2.45 chaser_return=6.23 explorer_return=58.26 vision=(0.000,0.000) new_fields=(81.8,72.8) distance=6.27 loss=(-0.001,3.727) kl=0.0005 +update=19500 collisions=2.48 chaser_return=6.39 explorer_return=63.14 vision=(0.000,0.000) new_fields=(82.8,76.0) distance=5.31 loss=(-0.001,3.639) kl=0.0002 +update=20000 collisions=2.17 chaser_return=6.30 explorer_return=40.08 vision=(0.000,0.000) new_fields=(82.0,60.6) distance=4.87 loss=(-0.001,4.015) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7575.8237598598935, "metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "task": "non_social", "unit_id": "non_social/seed_0007"} +=== exited 2026-07-02T11:50:20.933771+00:00 returncode=0 === diff --git a/logs/non_social_seed_0008_attempt_01.log b/logs/non_social_seed_0008_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..fc578c395a953a5b911cdc759e39f0ca66a4e9d5 --- /dev/null +++ b/logs/non_social_seed_0008_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:58.683758+00:00 task=non_social seed=8 attempt=1 === +update=0001 collisions=1.20 chaser_return=-3.33 explorer_return=-0.46 vision=(0.000,0.000) new_fields=(33.5,33.2) distance=5.33 loss=(-0.001,3.079) kl=0.0000 +update=0500 collisions=2.85 chaser_return=4.65 explorer_return=57.33 vision=(0.000,0.000) new_fields=(73.8,72.1) distance=5.43 loss=(-0.002,4.429) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=4.15 chaser_return=4.26 explorer_return=52.98 vision=(0.000,0.000) new_fields=(72.0,69.2) distance=5.34 loss=(-0.002,4.264) kl=0.0003 +update=1500 collisions=3.73 chaser_return=4.90 explorer_return=48.40 vision=(0.000,0.000) new_fields=(75.3,66.3) distance=5.17 loss=(-0.001,3.912) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=3.42 chaser_return=4.84 explorer_return=50.35 vision=(0.000,0.000) new_fields=(75.0,67.4) distance=5.65 loss=(-0.001,4.110) kl=0.0001 +update=2500 collisions=3.12 chaser_return=5.36 explorer_return=50.54 vision=(0.000,0.000) new_fields=(77.4,67.5) distance=4.78 loss=(-0.002,3.743) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=2.98 chaser_return=5.52 explorer_return=49.71 vision=(0.000,0.000) new_fields=(78.2,67.1) distance=4.63 loss=(-0.001,3.709) kl=0.0000 +update=3500 collisions=2.90 chaser_return=5.25 explorer_return=49.53 vision=(0.000,0.000) new_fields=(77.0,66.9) distance=4.52 loss=(-0.001,3.540) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=2.88 chaser_return=5.54 explorer_return=50.58 vision=(0.000,0.000) new_fields=(78.4,67.6) distance=5.88 loss=(-0.001,3.740) kl=0.0003 +update=4500 collisions=2.45 chaser_return=5.63 explorer_return=51.59 vision=(0.000,0.000) new_fields=(78.9,68.2) distance=4.98 loss=(-0.001,3.567) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=2.50 chaser_return=5.56 explorer_return=52.30 vision=(0.000,0.000) new_fields=(78.3,68.9) distance=5.54 loss=(-0.000,3.617) kl=0.0005 +update=5500 collisions=2.17 chaser_return=6.34 explorer_return=50.35 vision=(0.000,0.000) new_fields=(82.3,67.5) distance=4.98 loss=(-0.000,3.635) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.60 chaser_return=6.55 explorer_return=48.03 vision=(0.000,0.000) new_fields=(83.1,65.9) distance=5.21 loss=(-0.000,3.480) kl=0.0006 +update=6500 collisions=1.33 chaser_return=6.65 explorer_return=52.64 vision=(0.000,0.000) new_fields=(83.6,69.0) distance=5.85 loss=(-0.001,3.727) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=2.28 chaser_return=6.36 explorer_return=53.69 vision=(0.000,0.000) new_fields=(82.3,69.6) distance=4.77 loss=(-0.001,3.508) kl=0.0003 +update=7500 collisions=1.85 chaser_return=6.62 explorer_return=52.15 vision=(0.000,0.000) new_fields=(83.5,68.8) distance=4.47 loss=(-0.000,3.314) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=2.03 chaser_return=6.33 explorer_return=53.91 vision=(0.000,0.000) new_fields=(82.2,69.7) distance=5.06 loss=(-0.000,3.409) kl=0.0002 +update=8500 collisions=2.20 chaser_return=6.65 explorer_return=52.60 vision=(0.000,0.000) new_fields=(83.9,68.9) distance=5.03 loss=(-0.001,3.339) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=1.93 chaser_return=6.39 explorer_return=52.38 vision=(0.000,0.000) new_fields=(82.5,68.8) distance=5.01 loss=(-0.001,3.413) kl=0.0002 +update=9500 collisions=2.28 chaser_return=6.09 explorer_return=50.35 vision=(0.000,0.000) new_fields=(81.0,67.5) distance=5.02 loss=(-0.001,3.551) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=2.33 chaser_return=6.30 explorer_return=50.24 vision=(0.000,0.000) new_fields=(82.3,67.2) distance=5.40 loss=(-0.001,3.602) kl=0.0001 +update=10500 collisions=1.43 chaser_return=6.74 explorer_return=52.67 vision=(0.000,0.000) new_fields=(84.1,68.8) distance=6.27 loss=(-0.001,3.465) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=2.03 chaser_return=6.50 explorer_return=53.54 vision=(0.000,0.000) new_fields=(83.2,69.3) distance=5.15 loss=(-0.002,3.284) kl=0.0008 +update=11500 collisions=1.88 chaser_return=6.59 explorer_return=49.86 vision=(0.000,0.000) new_fields=(83.5,67.2) distance=5.48 loss=(-0.000,3.423) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=2.23 chaser_return=6.77 explorer_return=53.65 vision=(0.000,0.000) new_fields=(84.5,69.8) distance=5.38 loss=(-0.000,3.443) kl=0.0002 +update=12500 collisions=1.40 chaser_return=6.42 explorer_return=54.66 vision=(0.000,0.000) new_fields=(82.3,70.2) distance=5.47 loss=(-0.001,3.402) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=2.35 chaser_return=6.18 explorer_return=52.98 vision=(0.000,0.000) new_fields=(81.5,69.3) distance=5.13 loss=(-0.001,3.602) kl=0.0003 +update=13500 collisions=1.90 chaser_return=6.92 explorer_return=55.64 vision=(0.000,0.000) new_fields=(85.2,70.9) distance=5.89 loss=(-0.001,3.192) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=2.65 chaser_return=6.15 explorer_return=54.44 vision=(0.000,0.000) new_fields=(81.5,70.2) distance=4.89 loss=(-0.001,3.364) kl=0.0000 +update=14500 collisions=1.68 chaser_return=6.32 explorer_return=54.36 vision=(0.000,0.000) new_fields=(82.0,70.0) distance=5.32 loss=(-0.000,3.215) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=2.85 chaser_return=5.25 explorer_return=56.65 vision=(0.000,0.000) new_fields=(76.8,71.7) distance=5.54 loss=(-0.000,3.466) kl=0.0003 +update=15500 collisions=2.62 chaser_return=5.09 explorer_return=55.15 vision=(0.000,0.000) new_fields=(76.0,70.8) distance=5.29 loss=(-0.001,3.261) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=2.88 chaser_return=5.08 explorer_return=54.48 vision=(0.000,0.000) new_fields=(76.1,70.2) distance=4.54 loss=(-0.001,3.437) kl=-0.0001 +update=16500 collisions=2.78 chaser_return=5.18 explorer_return=57.40 vision=(0.000,0.000) new_fields=(76.6,72.2) distance=5.48 loss=(-0.000,3.312) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=3.08 chaser_return=5.32 explorer_return=56.35 vision=(0.000,0.000) new_fields=(77.3,71.5) distance=4.99 loss=(-0.001,3.467) kl=0.0007 +update=17500 collisions=2.88 chaser_return=4.70 explorer_return=58.34 vision=(0.000,0.000) new_fields=(74.2,72.6) distance=5.35 loss=(-0.001,3.363) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=2.95 chaser_return=4.89 explorer_return=57.17 vision=(0.000,0.000) new_fields=(75.3,72.0) distance=4.61 loss=(-0.001,3.421) kl=0.0007 +update=18500 collisions=2.38 chaser_return=4.96 explorer_return=57.40 vision=(0.000,0.000) new_fields=(75.2,72.0) distance=4.20 loss=(-0.001,3.389) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=2.05 chaser_return=6.01 explorer_return=55.60 vision=(0.000,0.000) new_fields=(80.5,70.9) distance=5.28 loss=(-0.001,3.723) kl=0.0003 +update=19500 collisions=2.75 chaser_return=5.87 explorer_return=56.58 vision=(0.000,0.000) new_fields=(80.1,71.5) distance=4.57 loss=(-0.000,3.602) kl=0.0002 +update=20000 collisions=1.98 chaser_return=6.15 explorer_return=52.49 vision=(0.000,0.000) new_fields=(81.3,68.7) distance=4.52 loss=(-0.001,4.003) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7577.7527450309135, "metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "task": "non_social", "unit_id": "non_social/seed_0008"} +=== exited 2026-07-02T11:50:20.938456+00:00 returncode=0 === diff --git a/logs/non_social_seed_0009_attempt_01.log b/logs/non_social_seed_0009_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..3f677235e692857ffd9a16520a79df03ddad1166 --- /dev/null +++ b/logs/non_social_seed_0009_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:58.685349+00:00 task=non_social seed=9 attempt=1 === +update=0001 collisions=1.35 chaser_return=-3.14 explorer_return=3.10 vision=(0.000,0.000) new_fields=(34.4,35.7) distance=4.75 loss=(-0.001,3.000) kl=0.0001 +update=0500 collisions=3.05 chaser_return=4.38 explorer_return=45.29 vision=(0.000,0.000) new_fields=(72.6,64.0) distance=5.09 loss=(-0.002,4.753) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=3.45 chaser_return=5.18 explorer_return=48.40 vision=(0.000,0.000) new_fields=(76.7,66.2) distance=4.81 loss=(-0.001,4.740) kl=0.0001 +update=1500 collisions=2.58 chaser_return=4.79 explorer_return=52.19 vision=(0.000,0.000) new_fields=(74.6,68.7) distance=6.12 loss=(-0.002,4.563) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=3.05 chaser_return=5.24 explorer_return=53.50 vision=(0.000,0.000) new_fields=(76.8,69.6) distance=4.50 loss=(-0.001,4.315) kl=0.0003 +update=2500 collisions=2.92 chaser_return=5.47 explorer_return=49.00 vision=(0.000,0.000) new_fields=(78.1,66.5) distance=5.13 loss=(-0.001,4.246) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=2.30 chaser_return=6.44 explorer_return=54.74 vision=(0.000,0.000) new_fields=(82.8,70.3) distance=5.01 loss=(-0.001,3.880) kl=0.0001 +update=3500 collisions=2.58 chaser_return=5.34 explorer_return=56.95 vision=(0.000,0.000) new_fields=(77.2,71.7) distance=5.21 loss=(-0.001,4.040) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=1.88 chaser_return=5.17 explorer_return=56.46 vision=(0.000,0.000) new_fields=(76.3,71.5) distance=4.51 loss=(-0.001,3.939) kl=0.0004 +update=4500 collisions=1.93 chaser_return=6.03 explorer_return=58.00 vision=(0.000,0.000) new_fields=(80.8,72.5) distance=5.79 loss=(-0.001,3.977) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=2.95 chaser_return=5.56 explorer_return=52.71 vision=(0.000,0.000) new_fields=(78.6,69.2) distance=5.54 loss=(-0.001,4.011) kl=-0.0001 +update=5500 collisions=2.40 chaser_return=5.75 explorer_return=57.40 vision=(0.000,0.000) new_fields=(79.2,72.1) distance=4.86 loss=(-0.001,3.808) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=1.55 chaser_return=5.84 explorer_return=58.67 vision=(0.000,0.000) new_fields=(79.8,72.9) distance=4.76 loss=(-0.001,3.803) kl=0.0003 +update=6500 collisions=1.73 chaser_return=4.94 explorer_return=53.95 vision=(0.000,0.000) new_fields=(75.2,69.8) distance=5.27 loss=(-0.001,4.034) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=2.95 chaser_return=6.07 explorer_return=59.58 vision=(0.000,0.000) new_fields=(81.0,73.5) distance=4.63 loss=(-0.001,3.707) kl=-0.0000 +update=7500 collisions=2.62 chaser_return=6.30 explorer_return=60.29 vision=(0.000,0.000) new_fields=(82.2,74.0) distance=5.25 loss=(-0.001,3.637) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=2.17 chaser_return=6.12 explorer_return=56.54 vision=(0.000,0.000) new_fields=(81.1,71.6) distance=5.69 loss=(-0.001,3.499) kl=0.0001 +update=8500 collisions=2.17 chaser_return=4.77 explorer_return=60.85 vision=(0.000,0.000) new_fields=(74.6,74.3) distance=5.21 loss=(-0.001,3.824) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=3.08 chaser_return=5.23 explorer_return=57.44 vision=(0.000,0.000) new_fields=(76.9,72.2) distance=5.10 loss=(-0.001,3.890) kl=0.0005 +update=9500 collisions=1.83 chaser_return=6.21 explorer_return=53.76 vision=(0.000,0.000) new_fields=(81.5,69.6) distance=3.76 loss=(-0.001,3.725) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=2.50 chaser_return=6.37 explorer_return=46.90 vision=(0.000,0.000) new_fields=(82.3,65.2) distance=4.34 loss=(-0.001,3.911) kl=0.0005 +update=10500 collisions=2.88 chaser_return=6.33 explorer_return=50.58 vision=(0.000,0.000) new_fields=(82.4,67.6) distance=4.59 loss=(-0.001,4.059) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=3.03 chaser_return=6.39 explorer_return=55.53 vision=(0.000,0.000) new_fields=(82.8,71.0) distance=4.56 loss=(-0.001,3.804) kl=0.0002 +update=11500 collisions=2.60 chaser_return=6.76 explorer_return=57.21 vision=(0.000,0.000) new_fields=(84.4,72.0) distance=5.05 loss=(-0.000,3.809) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=2.67 chaser_return=6.65 explorer_return=61.86 vision=(0.000,0.000) new_fields=(84.1,75.3) distance=3.69 loss=(-0.001,3.711) kl=0.0001 +update=12500 collisions=2.67 chaser_return=6.31 explorer_return=57.33 vision=(0.000,0.000) new_fields=(82.3,72.2) distance=5.71 loss=(-0.001,3.847) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=2.80 chaser_return=6.88 explorer_return=53.95 vision=(0.000,0.000) new_fields=(85.2,69.9) distance=5.24 loss=(-0.001,4.037) kl=0.0001 +update=13500 collisions=2.50 chaser_return=6.67 explorer_return=59.24 vision=(0.000,0.000) new_fields=(84.2,73.2) distance=5.39 loss=(-0.001,3.924) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=2.45 chaser_return=6.14 explorer_return=58.38 vision=(0.000,0.000) new_fields=(81.3,72.8) distance=5.25 loss=(-0.001,4.016) kl=0.0005 +update=14500 collisions=2.23 chaser_return=5.58 explorer_return=57.10 vision=(0.000,0.000) new_fields=(78.4,71.8) distance=4.94 loss=(-0.002,4.241) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=2.60 chaser_return=6.81 explorer_return=57.55 vision=(0.000,0.000) new_fields=(84.6,72.6) distance=4.95 loss=(-0.001,3.872) kl=0.0004 +update=15500 collisions=2.65 chaser_return=7.34 explorer_return=59.09 vision=(0.000,0.000) new_fields=(87.6,73.4) distance=5.42 loss=(-0.001,3.516) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=2.28 chaser_return=7.18 explorer_return=58.53 vision=(0.000,0.000) new_fields=(86.7,73.0) distance=5.84 loss=(-0.001,3.693) kl=0.0005 +update=16500 collisions=2.40 chaser_return=6.90 explorer_return=57.14 vision=(0.000,0.000) new_fields=(85.0,72.0) distance=5.23 loss=(-0.001,3.883) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=2.03 chaser_return=6.84 explorer_return=60.89 vision=(0.000,0.000) new_fields=(84.8,74.5) distance=5.42 loss=(-0.000,3.668) kl=0.0001 +update=17500 collisions=3.33 chaser_return=6.56 explorer_return=57.21 vision=(0.000,0.000) new_fields=(83.6,72.1) distance=5.64 loss=(-0.001,3.709) kl=-0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=2.05 chaser_return=6.86 explorer_return=64.71 vision=(0.000,0.000) new_fields=(84.9,77.2) distance=5.80 loss=(-0.001,3.714) kl=0.0002 +update=18500 collisions=2.42 chaser_return=6.88 explorer_return=47.05 vision=(0.000,0.000) new_fields=(85.3,65.2) distance=5.64 loss=(-0.001,3.949) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=1.93 chaser_return=6.77 explorer_return=57.85 vision=(0.000,0.000) new_fields=(84.3,72.3) distance=5.45 loss=(-0.001,3.705) kl=0.0000 +update=19500 collisions=2.05 chaser_return=6.73 explorer_return=59.69 vision=(0.000,0.000) new_fields=(84.2,73.7) distance=5.33 loss=(-0.000,3.773) kl=-0.0001 +update=20000 collisions=2.73 chaser_return=6.34 explorer_return=47.46 vision=(0.000,0.000) new_fields=(82.6,65.7) distance=5.68 loss=(-0.001,3.992) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7570.80278072902, "metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "task": "non_social", "unit_id": "non_social/seed_0009"} +=== exited 2026-07-02T11:50:20.942911+00:00 returncode=0 === diff --git a/logs/social_seed_0000_attempt_01.log b/logs/social_seed_0000_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..8899b56ad8b3549be22faf183216454ed1f5d061 --- /dev/null +++ b/logs/social_seed_0000_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.521136+00:00 task=social seed=0 attempt=1 === +update=0001 collisions=1.55 chaser_return=-1.18 explorer_return=-0.41 vision=(0.343,0.343) new_fields=(35.9,33.9) distance=5.38 loss=(-0.001,3.128) kl=-0.0000 +update=0500 collisions=13.38 chaser_return=15.03 explorer_return=35.67 vision=(0.875,0.875) new_fields=(54.5,64.4) distance=2.02 loss=(-0.006,5.704) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=13.23 chaser_return=14.47 explorer_return=33.01 vision=(0.868,0.868) new_fields=(52.7,63.3) distance=2.46 loss=(-0.008,5.706) kl=0.0012 +update=1500 collisions=14.07 chaser_return=15.30 explorer_return=35.14 vision=(0.844,0.844) new_fields=(52.6,65.2) distance=2.23 loss=(-0.006,5.844) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=15.45 chaser_return=16.17 explorer_return=29.43 vision=(0.933,0.933) new_fields=(49.6,62.4) distance=2.19 loss=(-0.005,5.605) kl=0.0012 +update=2500 collisions=16.85 chaser_return=17.93 explorer_return=26.55 vision=(0.924,0.924) new_fields=(50.5,60.9) distance=1.91 loss=(-0.005,5.639) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=15.62 chaser_return=16.34 explorer_return=31.66 vision=(0.934,0.934) new_fields=(49.5,63.9) distance=1.93 loss=(-0.005,5.336) kl=0.0006 +update=3500 collisions=20.50 chaser_return=21.50 explorer_return=25.25 vision=(0.940,0.940) new_fields=(49.3,62.2) distance=1.81 loss=(-0.005,5.557) kl=0.0015 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=21.05 chaser_return=22.48 explorer_return=22.31 vision=(0.954,0.954) new_fields=(50.6,60.0) distance=1.64 loss=(-0.005,5.912) kl=0.0010 +update=4500 collisions=16.83 chaser_return=17.33 explorer_return=28.96 vision=(0.930,0.930) new_fields=(48.0,62.9) distance=1.88 loss=(-0.006,5.590) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=20.12 chaser_return=21.21 explorer_return=28.25 vision=(0.931,0.931) new_fields=(50.1,63.9) distance=1.69 loss=(-0.004,5.518) kl=0.0004 +update=5500 collisions=16.23 chaser_return=16.47 explorer_return=20.86 vision=(0.930,0.930) new_fields=(47.4,57.3) distance=1.66 loss=(-0.004,5.827) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=20.88 chaser_return=21.59 explorer_return=23.86 vision=(0.955,0.955) new_fields=(48.8,62.4) distance=1.52 loss=(-0.003,5.077) kl=0.0006 +update=6500 collisions=18.62 chaser_return=19.72 explorer_return=27.16 vision=(0.949,0.949) new_fields=(50.3,63.4) distance=1.64 loss=(-0.003,5.263) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=18.43 chaser_return=19.00 explorer_return=26.48 vision=(0.932,0.932) new_fields=(47.8,61.7) distance=1.77 loss=(-0.006,5.619) kl=0.0003 +update=7500 collisions=21.52 chaser_return=22.36 explorer_return=25.56 vision=(0.940,0.940) new_fields=(48.4,63.3) distance=1.66 loss=(-0.004,5.634) kl=-0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=17.77 chaser_return=18.38 explorer_return=32.91 vision=(0.947,0.947) new_fields=(48.4,67.3) distance=1.71 loss=(-0.003,5.241) kl=0.0005 +update=8500 collisions=21.77 chaser_return=22.35 explorer_return=22.14 vision=(0.944,0.944) new_fields=(46.4,60.8) distance=1.55 loss=(-0.004,5.629) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=21.98 chaser_return=23.22 explorer_return=23.76 vision=(0.946,0.946) new_fields=(49.9,61.9) distance=1.76 loss=(-0.005,5.505) kl=0.0005 +update=9500 collisions=20.20 chaser_return=20.92 explorer_return=19.51 vision=(0.960,0.960) new_fields=(47.4,58.1) distance=1.76 loss=(-0.006,5.600) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=20.05 chaser_return=21.00 explorer_return=31.66 vision=(0.919,0.919) new_fields=(48.2,65.7) distance=2.10 loss=(-0.004,5.301) kl=0.0008 +update=10500 collisions=17.05 chaser_return=18.19 explorer_return=31.70 vision=(0.952,0.952) new_fields=(50.2,64.2) distance=1.99 loss=(-0.003,5.351) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=19.35 chaser_return=20.85 explorer_return=25.49 vision=(0.929,0.929) new_fields=(52.7,62.2) distance=1.61 loss=(-0.003,5.082) kl=0.0003 +update=11500 collisions=24.95 chaser_return=26.32 explorer_return=20.96 vision=(0.940,0.940) new_fields=(49.4,61.0) distance=1.45 loss=(-0.004,5.243) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=20.05 chaser_return=21.22 explorer_return=29.38 vision=(0.943,0.943) new_fields=(51.3,65.8) distance=1.68 loss=(-0.003,4.689) kl=0.0005 +update=12500 collisions=17.50 chaser_return=18.93 explorer_return=25.74 vision=(0.904,0.904) new_fields=(51.9,60.3) distance=1.66 loss=(-0.005,5.803) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=20.23 chaser_return=21.15 explorer_return=25.20 vision=(0.954,0.954) new_fields=(49.1,62.7) distance=1.55 loss=(-0.006,5.374) kl=0.0004 +update=13500 collisions=18.08 chaser_return=19.56 explorer_return=26.20 vision=(0.923,0.923) new_fields=(51.6,61.2) distance=1.71 loss=(-0.005,5.471) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=19.35 chaser_return=20.41 explorer_return=23.16 vision=(0.955,0.955) new_fields=(48.9,59.9) distance=1.65 loss=(-0.004,5.519) kl=0.0006 +update=14500 collisions=20.73 chaser_return=21.12 explorer_return=18.20 vision=(0.912,0.912) new_fields=(46.4,57.0) distance=1.97 loss=(-0.006,5.378) kl=0.0015 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=17.10 chaser_return=18.27 explorer_return=24.59 vision=(0.946,0.946) new_fields=(51.8,60.2) distance=1.70 loss=(-0.004,4.840) kl=0.0013 +update=15500 collisions=17.73 chaser_return=18.79 explorer_return=23.94 vision=(0.929,0.929) new_fields=(50.7,59.8) distance=1.70 loss=(-0.005,4.998) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=18.58 chaser_return=19.77 explorer_return=23.77 vision=(0.910,0.910) new_fields=(51.1,58.9) distance=1.97 loss=(-0.003,5.528) kl=0.0001 +update=16500 collisions=23.68 chaser_return=24.70 explorer_return=14.51 vision=(0.934,0.934) new_fields=(47.8,56.2) distance=1.58 loss=(-0.006,5.932) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=20.43 chaser_return=21.99 explorer_return=21.31 vision=(0.959,0.959) new_fields=(52.9,60.1) distance=1.48 loss=(-0.004,5.102) kl=0.0006 +update=17500 collisions=18.43 chaser_return=19.79 explorer_return=20.18 vision=(0.963,0.963) new_fields=(51.8,57.0) distance=1.54 loss=(-0.002,5.649) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=18.98 chaser_return=20.29 explorer_return=19.71 vision=(0.953,0.953) new_fields=(50.6,57.0) distance=1.57 loss=(-0.003,4.677) kl=-0.0003 +update=18500 collisions=19.60 chaser_return=20.53 explorer_return=22.33 vision=(0.970,0.970) new_fields=(50.8,60.7) distance=1.52 loss=(-0.003,5.175) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=19.85 chaser_return=21.41 explorer_return=21.23 vision=(0.938,0.938) new_fields=(52.0,58.1) distance=1.78 loss=(-0.002,4.832) kl=0.0003 +update=19500 collisions=17.95 chaser_return=18.80 explorer_return=24.95 vision=(0.943,0.943) new_fields=(50.0,61.0) distance=1.59 loss=(-0.004,5.202) kl=0.0002 +update=20000 collisions=21.45 chaser_return=23.26 explorer_return=31.04 vision=(0.949,0.949) new_fields=(52.9,65.6) distance=1.66 loss=(-0.004,4.788) kl=0.0028 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7633.112761525088, "metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "task": "social", "unit_id": "social/seed_0000"} +=== exited 2026-07-02T09:43:38.591318+00:00 returncode=0 === diff --git a/logs/social_seed_0001_attempt_01.log b/logs/social_seed_0001_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..45194f494b7d07d28d1295a1b28961339b1b762e --- /dev/null +++ b/logs/social_seed_0001_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.522668+00:00 task=social seed=1 attempt=1 === +update=0001 collisions=1.08 chaser_return=-1.94 explorer_return=3.50 vision=(0.320,0.320) new_fields=(34.5,36.2) distance=5.04 loss=(-0.001,2.967) kl=-0.0000 +update=0500 collisions=10.82 chaser_return=11.62 explorer_return=33.88 vision=(0.882,0.882) new_fields=(51.0,62.0) distance=2.11 loss=(-0.006,5.540) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=12.32 chaser_return=13.37 explorer_return=29.60 vision=(0.886,0.886) new_fields=(51.1,59.9) distance=1.95 loss=(-0.007,5.394) kl=0.0010 +update=1500 collisions=18.95 chaser_return=20.12 explorer_return=25.50 vision=(0.840,0.840) new_fields=(48.8,59.4) distance=2.35 loss=(-0.002,5.995) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=13.00 chaser_return=13.99 explorer_return=35.08 vision=(0.858,0.858) new_fields=(51.5,64.4) distance=2.15 loss=(-0.005,5.452) kl=0.0006 +update=2500 collisions=14.62 chaser_return=15.85 explorer_return=38.39 vision=(0.842,0.842) new_fields=(52.2,67.6) distance=2.44 loss=(-0.004,5.462) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=15.00 chaser_return=16.84 explorer_return=40.11 vision=(0.875,0.875) new_fields=(55.0,69.4) distance=1.66 loss=(-0.004,5.187) kl=0.0004 +update=3500 collisions=17.12 chaser_return=18.35 explorer_return=30.01 vision=(0.924,0.924) new_fields=(51.5,63.0) distance=1.99 loss=(-0.006,5.489) kl=0.0013 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=13.88 chaser_return=14.88 explorer_return=30.36 vision=(0.899,0.899) new_fields=(51.5,61.6) distance=1.93 loss=(-0.006,5.385) kl=0.0006 +update=4500 collisions=16.40 chaser_return=17.02 explorer_return=27.64 vision=(0.900,0.900) new_fields=(48.6,61.1) distance=1.71 loss=(-0.005,5.530) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=17.12 chaser_return=17.93 explorer_return=29.30 vision=(0.891,0.891) new_fields=(49.4,62.9) distance=1.77 loss=(-0.006,5.748) kl=0.0005 +update=5500 collisions=17.27 chaser_return=18.28 explorer_return=28.96 vision=(0.916,0.916) new_fields=(49.4,62.3) distance=2.05 loss=(-0.007,5.578) kl=0.0014 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=16.18 chaser_return=17.09 explorer_return=29.21 vision=(0.945,0.945) new_fields=(50.2,61.4) distance=1.69 loss=(-0.005,5.509) kl=0.0007 +update=6500 collisions=18.65 chaser_return=19.88 explorer_return=28.69 vision=(0.934,0.934) new_fields=(50.1,63.0) distance=1.69 loss=(-0.004,5.050) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=18.12 chaser_return=19.23 explorer_return=24.83 vision=(0.947,0.947) new_fields=(50.8,59.3) distance=1.47 loss=(-0.005,5.500) kl=0.0010 +update=7500 collisions=14.95 chaser_return=16.35 explorer_return=31.18 vision=(0.942,0.942) new_fields=(53.8,63.2) distance=1.81 loss=(-0.003,5.200) kl=-0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=21.08 chaser_return=22.54 explorer_return=27.33 vision=(0.953,0.953) new_fields=(52.5,64.6) distance=1.66 loss=(-0.002,5.103) kl=0.0002 +update=8500 collisions=23.15 chaser_return=24.65 explorer_return=20.10 vision=(0.942,0.942) new_fields=(50.9,59.7) distance=1.72 loss=(-0.005,5.367) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=19.08 chaser_return=18.64 explorer_return=18.46 vision=(0.939,0.939) new_fields=(43.0,57.1) distance=1.62 loss=(-0.005,5.450) kl=0.0004 +update=9500 collisions=24.18 chaser_return=24.82 explorer_return=10.48 vision=(0.945,0.945) new_fields=(45.9,53.7) distance=1.99 loss=(-0.003,5.661) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=19.08 chaser_return=19.90 explorer_return=20.08 vision=(0.940,0.940) new_fields=(49.4,58.6) distance=1.61 loss=(-0.004,5.291) kl=0.0007 +update=10500 collisions=23.60 chaser_return=25.38 explorer_return=24.52 vision=(0.954,0.954) new_fields=(52.6,64.8) distance=1.67 loss=(-0.004,4.451) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=20.75 chaser_return=22.15 explorer_return=21.68 vision=(0.941,0.941) new_fields=(51.9,59.5) distance=1.69 loss=(-0.005,5.132) kl=0.0006 +update=11500 collisions=24.48 chaser_return=26.40 explorer_return=19.29 vision=(0.946,0.946) new_fields=(53.0,61.4) distance=1.70 loss=(-0.004,4.278) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=25.12 chaser_return=26.09 explorer_return=14.09 vision=(0.941,0.941) new_fields=(45.2,55.6) distance=1.57 loss=(-0.005,5.136) kl=0.0003 +update=12500 collisions=25.00 chaser_return=26.43 explorer_return=18.61 vision=(0.958,0.958) new_fields=(49.6,60.0) distance=1.65 loss=(-0.005,5.023) kl=0.0018 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=16.98 chaser_return=17.30 explorer_return=18.76 vision=(0.916,0.916) new_fields=(46.1,54.7) distance=1.80 loss=(-0.004,5.845) kl=0.0006 +update=13500 collisions=23.68 chaser_return=23.58 explorer_return=14.25 vision=(0.899,0.899) new_fields=(41.4,54.0) distance=2.19 loss=(-0.004,6.881) kl=0.0013 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=20.45 chaser_return=21.04 explorer_return=21.56 vision=(0.942,0.942) new_fields=(46.6,59.7) distance=1.82 loss=(-0.004,6.012) kl=0.0006 +update=14500 collisions=20.18 chaser_return=21.29 explorer_return=25.68 vision=(0.950,0.950) new_fields=(49.2,61.6) distance=1.64 loss=(-0.005,5.376) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=20.58 chaser_return=22.01 explorer_return=20.19 vision=(0.961,0.961) new_fields=(50.5,57.2) distance=1.80 loss=(-0.005,5.727) kl=0.0016 +update=15500 collisions=18.43 chaser_return=19.35 explorer_return=21.41 vision=(0.909,0.909) new_fields=(50.2,58.9) distance=1.76 loss=(-0.005,5.373) kl=0.0013 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=18.23 chaser_return=19.38 explorer_return=19.41 vision=(0.952,0.952) new_fields=(50.3,57.0) distance=1.73 loss=(-0.004,5.053) kl=0.0007 +update=16500 collisions=17.15 chaser_return=18.02 explorer_return=22.27 vision=(0.935,0.935) new_fields=(49.4,57.2) distance=1.61 loss=(-0.004,5.534) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=19.30 chaser_return=20.44 explorer_return=25.29 vision=(0.952,0.952) new_fields=(48.6,61.2) distance=1.56 loss=(-0.004,4.820) kl=-0.0001 +update=17500 collisions=15.70 chaser_return=17.64 explorer_return=26.94 vision=(0.959,0.959) new_fields=(55.2,60.1) distance=1.57 loss=(-0.003,5.384) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=19.10 chaser_return=20.42 explorer_return=20.29 vision=(0.918,0.918) new_fields=(50.9,58.2) distance=1.51 loss=(-0.002,5.023) kl=0.0002 +update=18500 collisions=14.98 chaser_return=16.10 explorer_return=28.05 vision=(0.936,0.936) new_fields=(50.9,60.4) distance=2.05 loss=(-0.005,5.323) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=18.62 chaser_return=20.25 explorer_return=19.96 vision=(0.938,0.938) new_fields=(51.3,57.3) distance=1.63 loss=(-0.004,5.188) kl=0.0009 +update=19500 collisions=18.80 chaser_return=19.55 explorer_return=23.18 vision=(0.946,0.946) new_fields=(48.9,59.8) distance=1.64 loss=(-0.005,5.439) kl=0.0011 +update=20000 collisions=22.62 chaser_return=23.43 explorer_return=12.71 vision=(0.930,0.930) new_fields=(46.2,52.8) distance=1.79 loss=(-0.004,5.151) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7644.622009031009, "metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "task": "social", "unit_id": "social/seed_0001"} +=== exited 2026-07-02T09:43:48.618767+00:00 returncode=0 === diff --git a/logs/social_seed_0002_attempt_01.log b/logs/social_seed_0002_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..6fd042fa270e65c546f59c7de6e76a2ebbf1c830 --- /dev/null +++ b/logs/social_seed_0002_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.523612+00:00 task=social seed=2 attempt=1 === +update=0001 collisions=1.77 chaser_return=-1.36 explorer_return=4.43 vision=(0.351,0.351) new_fields=(33.7,37.0) distance=5.46 loss=(-0.001,3.191) kl=-0.0001 +update=0500 collisions=13.62 chaser_return=15.27 explorer_return=31.39 vision=(0.913,0.913) new_fields=(54.1,61.9) distance=2.02 loss=(-0.006,5.733) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=12.88 chaser_return=13.40 explorer_return=33.38 vision=(0.864,0.864) new_fields=(49.1,63.1) distance=2.16 loss=(-0.006,5.659) kl=0.0003 +update=1500 collisions=17.75 chaser_return=18.26 explorer_return=23.36 vision=(0.858,0.858) new_fields=(46.0,57.4) distance=2.22 loss=(-0.003,6.222) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=16.00 chaser_return=17.11 explorer_return=32.00 vision=(0.925,0.925) new_fields=(51.2,64.2) distance=1.80 loss=(-0.005,5.407) kl=0.0007 +update=2500 collisions=17.95 chaser_return=18.81 explorer_return=29.26 vision=(0.933,0.933) new_fields=(49.4,63.5) distance=1.71 loss=(-0.007,5.811) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=16.08 chaser_return=17.28 explorer_return=27.65 vision=(0.920,0.920) new_fields=(50.7,60.1) distance=1.92 loss=(-0.003,5.428) kl=0.0005 +update=3500 collisions=17.30 chaser_return=18.31 explorer_return=30.00 vision=(0.931,0.931) new_fields=(51.2,63.0) distance=1.75 loss=(-0.005,4.935) kl=0.0012 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=14.70 chaser_return=15.64 explorer_return=28.75 vision=(0.926,0.926) new_fields=(50.6,61.3) distance=1.90 loss=(-0.007,5.421) kl=0.0012 +update=4500 collisions=16.33 chaser_return=17.15 explorer_return=28.99 vision=(0.953,0.953) new_fields=(50.0,62.2) distance=1.75 loss=(-0.004,5.477) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=15.65 chaser_return=16.39 explorer_return=29.44 vision=(0.920,0.920) new_fields=(49.9,61.9) distance=1.92 loss=(-0.004,5.663) kl=0.0002 +update=5500 collisions=18.98 chaser_return=20.03 explorer_return=26.16 vision=(0.940,0.940) new_fields=(49.4,61.7) distance=1.63 loss=(-0.004,5.522) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=18.52 chaser_return=19.91 explorer_return=28.38 vision=(0.935,0.935) new_fields=(51.8,62.0) distance=1.78 loss=(-0.004,5.418) kl=0.0013 +update=6500 collisions=18.98 chaser_return=20.61 explorer_return=28.56 vision=(0.959,0.959) new_fields=(52.0,62.8) distance=1.77 loss=(-0.005,5.486) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=18.45 chaser_return=20.15 explorer_return=25.38 vision=(0.939,0.939) new_fields=(52.7,60.4) distance=1.61 loss=(-0.004,5.538) kl=0.0002 +update=7500 collisions=17.62 chaser_return=19.13 explorer_return=25.86 vision=(0.910,0.910) new_fields=(52.7,60.3) distance=1.91 loss=(-0.005,5.519) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=19.00 chaser_return=20.10 explorer_return=27.65 vision=(0.905,0.905) new_fields=(49.9,61.9) distance=1.72 loss=(-0.004,5.314) kl=0.0009 +update=8500 collisions=16.90 chaser_return=17.89 explorer_return=29.19 vision=(0.924,0.924) new_fields=(50.6,62.4) distance=1.69 loss=(-0.006,5.260) kl=0.0020 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=18.25 chaser_return=19.01 explorer_return=25.62 vision=(0.889,0.889) new_fields=(48.4,61.0) distance=1.78 loss=(-0.005,5.587) kl=0.0002 +update=9500 collisions=20.48 chaser_return=21.45 explorer_return=30.62 vision=(0.917,0.917) new_fields=(48.3,65.8) distance=1.77 loss=(-0.004,4.956) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=18.60 chaser_return=19.53 explorer_return=26.39 vision=(0.941,0.941) new_fields=(49.5,61.5) distance=1.78 loss=(-0.005,5.054) kl=0.0009 +update=10500 collisions=21.20 chaser_return=21.32 explorer_return=23.14 vision=(0.926,0.926) new_fields=(44.2,60.2) distance=1.72 loss=(-0.006,5.529) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=21.00 chaser_return=21.79 explorer_return=25.79 vision=(0.923,0.923) new_fields=(47.7,63.3) distance=1.73 loss=(-0.006,5.579) kl=-0.0002 +update=11500 collisions=20.48 chaser_return=22.33 explorer_return=28.60 vision=(0.945,0.945) new_fields=(53.2,65.0) distance=1.74 loss=(-0.005,4.895) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=19.62 chaser_return=20.70 explorer_return=20.62 vision=(0.915,0.915) new_fields=(49.9,58.0) distance=1.68 loss=(-0.003,5.800) kl=0.0005 +update=12500 collisions=19.58 chaser_return=20.77 explorer_return=21.89 vision=(0.958,0.958) new_fields=(49.8,58.9) distance=1.42 loss=(-0.004,5.128) kl=0.0015 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=20.93 chaser_return=22.59 explorer_return=23.69 vision=(0.930,0.930) new_fields=(53.0,61.8) distance=1.71 loss=(-0.004,5.010) kl=0.0006 +update=13500 collisions=20.80 chaser_return=21.57 explorer_return=21.24 vision=(0.934,0.934) new_fields=(47.9,58.9) distance=1.74 loss=(-0.005,5.529) kl=-0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=20.58 chaser_return=21.48 explorer_return=22.44 vision=(0.943,0.943) new_fields=(48.5,59.8) distance=1.82 loss=(-0.006,5.330) kl=0.0012 +update=14500 collisions=18.68 chaser_return=19.69 explorer_return=20.39 vision=(0.939,0.939) new_fields=(49.4,56.9) distance=1.82 loss=(-0.004,4.297) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=17.77 chaser_return=19.23 explorer_return=27.59 vision=(0.923,0.923) new_fields=(52.3,61.8) distance=1.95 loss=(-0.005,5.129) kl=0.0008 +update=15500 collisions=18.85 chaser_return=20.09 explorer_return=26.49 vision=(0.941,0.941) new_fields=(50.5,61.5) distance=1.80 loss=(-0.004,5.276) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=16.60 chaser_return=17.45 explorer_return=27.20 vision=(0.928,0.928) new_fields=(49.4,60.4) distance=1.57 loss=(-0.004,4.995) kl=0.0008 +update=16500 collisions=20.38 chaser_return=21.55 explorer_return=18.19 vision=(0.952,0.952) new_fields=(50.8,58.0) distance=1.53 loss=(-0.003,4.908) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=31.12 chaser_return=32.67 explorer_return=9.59 vision=(0.918,0.918) new_fields=(46.0,53.4) distance=1.52 loss=(-0.004,6.142) kl=0.0004 +update=17500 collisions=20.68 chaser_return=22.20 explorer_return=20.66 vision=(0.942,0.942) new_fields=(52.6,60.0) distance=1.58 loss=(-0.004,5.418) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=17.38 chaser_return=18.61 explorer_return=22.95 vision=(0.931,0.931) new_fields=(51.8,58.6) distance=1.69 loss=(-0.005,5.352) kl=0.0002 +update=18500 collisions=20.33 chaser_return=21.49 explorer_return=18.62 vision=(0.941,0.941) new_fields=(50.1,57.3) distance=1.71 loss=(-0.002,5.454) kl=0.0012 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=16.10 chaser_return=17.36 explorer_return=24.26 vision=(0.931,0.931) new_fields=(52.0,59.3) distance=1.59 loss=(-0.004,5.270) kl=0.0014 +update=19500 collisions=20.08 chaser_return=20.62 explorer_return=17.93 vision=(0.900,0.900) new_fields=(46.5,55.7) distance=1.88 loss=(-0.005,5.528) kl=0.0003 +update=20000 collisions=17.62 chaser_return=18.49 explorer_return=26.84 vision=(0.929,0.929) new_fields=(49.3,61.0) distance=1.66 loss=(-0.005,5.285) kl=0.0014 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7640.536813838058, "metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "task": "social", "unit_id": "social/seed_0002"} +=== exited 2026-07-02T09:43:48.626929+00:00 returncode=0 === diff --git a/logs/social_seed_0003_attempt_01.log b/logs/social_seed_0003_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..8900b9c8aef3cb1e6b76a8d0d383ede42187496a --- /dev/null +++ b/logs/social_seed_0003_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.524464+00:00 task=social seed=3 attempt=1 === +update=0001 collisions=2.25 chaser_return=-1.33 explorer_return=-1.62 vision=(0.335,0.335) new_fields=(31.4,33.2) distance=4.66 loss=(-0.001,3.365) kl=0.0000 +update=0500 collisions=7.75 chaser_return=9.81 explorer_return=39.61 vision=(0.486,0.486) new_fields=(57.7,63.9) distance=4.32 loss=(-0.006,5.427) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=9.12 chaser_return=9.90 explorer_return=33.23 vision=(0.628,0.628) new_fields=(50.7,60.5) distance=3.20 loss=(-0.008,5.689) kl=0.0009 +update=1500 collisions=14.20 chaser_return=15.11 explorer_return=30.39 vision=(0.712,0.712) new_fields=(49.8,60.7) distance=3.09 loss=(-0.008,5.967) kl=0.0017 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=10.93 chaser_return=10.56 explorer_return=27.79 vision=(0.750,0.750) new_fields=(45.0,57.4) distance=2.43 loss=(-0.007,5.428) kl=0.0005 +update=2500 collisions=14.30 chaser_return=15.41 explorer_return=33.41 vision=(0.877,0.877) new_fields=(51.8,63.2) distance=1.97 loss=(-0.005,5.269) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=14.50 chaser_return=14.96 explorer_return=31.29 vision=(0.872,0.872) new_fields=(48.7,62.7) distance=2.60 loss=(-0.007,5.670) kl=0.0006 +update=3500 collisions=16.88 chaser_return=18.27 explorer_return=29.99 vision=(0.933,0.933) new_fields=(52.5,62.2) distance=2.05 loss=(-0.006,5.498) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=16.10 chaser_return=17.56 explorer_return=28.27 vision=(0.921,0.921) new_fields=(53.5,61.1) distance=2.00 loss=(-0.005,5.506) kl=0.0008 +update=4500 collisions=17.85 chaser_return=18.94 explorer_return=28.90 vision=(0.929,0.929) new_fields=(51.2,62.0) distance=1.85 loss=(-0.005,5.304) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=18.10 chaser_return=19.14 explorer_return=26.71 vision=(0.935,0.935) new_fields=(50.2,61.6) distance=2.15 loss=(-0.006,5.127) kl=0.0009 +update=5500 collisions=27.62 chaser_return=28.64 explorer_return=14.11 vision=(0.922,0.922) new_fields=(44.5,55.9) distance=1.75 loss=(-0.004,6.265) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=17.93 chaser_return=18.61 explorer_return=25.60 vision=(0.934,0.934) new_fields=(48.2,61.0) distance=1.95 loss=(-0.007,5.765) kl=0.0009 +update=6500 collisions=18.23 chaser_return=18.96 explorer_return=25.68 vision=(0.947,0.947) new_fields=(48.1,60.9) distance=1.87 loss=(-0.006,5.595) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=18.50 chaser_return=19.01 explorer_return=25.61 vision=(0.937,0.937) new_fields=(47.1,60.7) distance=1.81 loss=(-0.006,5.844) kl=0.0009 +update=7500 collisions=20.65 chaser_return=21.87 explorer_return=24.39 vision=(0.939,0.939) new_fields=(49.7,61.5) distance=1.63 loss=(-0.006,5.313) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=18.75 chaser_return=20.29 explorer_return=29.73 vision=(0.895,0.895) new_fields=(52.6,63.7) distance=1.94 loss=(-0.005,5.260) kl=0.0006 +update=8500 collisions=17.33 chaser_return=18.82 explorer_return=26.88 vision=(0.901,0.901) new_fields=(53.4,61.0) distance=1.99 loss=(-0.005,5.187) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=23.55 chaser_return=25.10 explorer_return=27.02 vision=(0.943,0.943) new_fields=(50.5,65.3) distance=1.62 loss=(-0.004,4.676) kl=0.0011 +update=9500 collisions=18.52 chaser_return=18.94 explorer_return=23.65 vision=(0.883,0.883) new_fields=(46.2,60.1) distance=1.90 loss=(-0.005,5.649) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=17.68 chaser_return=17.84 explorer_return=23.77 vision=(0.940,0.940) new_fields=(45.7,59.8) distance=2.10 loss=(-0.007,5.642) kl=0.0009 +update=10500 collisions=16.62 chaser_return=17.66 explorer_return=23.59 vision=(0.919,0.919) new_fields=(50.0,58.5) distance=1.86 loss=(-0.006,5.099) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=20.35 chaser_return=21.21 explorer_return=23.30 vision=(0.943,0.943) new_fields=(48.4,60.0) distance=1.65 loss=(-0.005,5.237) kl=0.0006 +update=11500 collisions=15.95 chaser_return=16.83 explorer_return=27.15 vision=(0.935,0.935) new_fields=(49.8,61.0) distance=1.67 loss=(-0.003,5.660) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=19.77 chaser_return=20.86 explorer_return=19.58 vision=(0.940,0.940) new_fields=(49.3,56.7) distance=1.87 loss=(-0.003,5.213) kl=0.0003 +update=12500 collisions=19.35 chaser_return=20.05 explorer_return=20.99 vision=(0.935,0.935) new_fields=(48.2,58.4) distance=1.66 loss=(-0.004,5.335) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=19.25 chaser_return=19.34 explorer_return=18.49 vision=(0.945,0.945) new_fields=(44.5,56.5) distance=2.06 loss=(-0.006,5.764) kl=0.0005 +update=13500 collisions=23.23 chaser_return=24.17 explorer_return=16.39 vision=(0.938,0.938) new_fields=(49.0,58.2) distance=1.79 loss=(-0.005,5.262) kl=0.0013 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=20.20 chaser_return=21.58 explorer_return=20.90 vision=(0.945,0.945) new_fields=(51.0,59.2) distance=1.94 loss=(-0.004,4.611) kl=0.0011 +update=14500 collisions=16.12 chaser_return=17.18 explorer_return=31.23 vision=(0.903,0.903) new_fields=(51.6,63.5) distance=2.14 loss=(-0.006,5.432) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=16.05 chaser_return=16.87 explorer_return=22.45 vision=(0.931,0.931) new_fields=(49.2,57.8) distance=1.76 loss=(-0.004,5.438) kl=0.0005 +update=15500 collisions=20.05 chaser_return=20.75 explorer_return=20.41 vision=(0.947,0.947) new_fields=(49.0,59.2) distance=1.56 loss=(-0.005,5.328) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=12.62 chaser_return=12.98 explorer_return=29.41 vision=(0.926,0.926) new_fields=(48.0,59.9) distance=1.76 loss=(-0.004,6.173) kl=0.0011 +update=16500 collisions=20.08 chaser_return=19.93 explorer_return=14.32 vision=(0.928,0.928) new_fields=(43.2,53.1) distance=1.80 loss=(-0.005,5.542) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=20.80 chaser_return=21.68 explorer_return=18.69 vision=(0.940,0.940) new_fields=(48.4,57.7) distance=1.56 loss=(-0.004,5.817) kl=0.0008 +update=17500 collisions=15.78 chaser_return=16.62 explorer_return=28.48 vision=(0.920,0.920) new_fields=(49.9,61.6) distance=1.89 loss=(-0.005,5.507) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=19.35 chaser_return=20.32 explorer_return=21.25 vision=(0.959,0.959) new_fields=(49.2,59.3) distance=1.61 loss=(-0.006,5.306) kl=0.0008 +update=18500 collisions=23.23 chaser_return=24.16 explorer_return=19.05 vision=(0.934,0.934) new_fields=(48.2,58.4) distance=1.72 loss=(-0.005,5.771) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=21.02 chaser_return=22.85 explorer_return=26.49 vision=(0.946,0.946) new_fields=(52.6,63.4) distance=1.78 loss=(-0.004,4.452) kl=-0.0002 +update=19500 collisions=18.98 chaser_return=19.46 explorer_return=19.11 vision=(0.929,0.929) new_fields=(46.8,56.8) distance=1.70 loss=(-0.004,5.291) kl=0.0002 +update=20000 collisions=20.20 chaser_return=21.88 explorer_return=22.89 vision=(0.930,0.930) new_fields=(52.5,59.3) distance=1.49 loss=(-0.003,5.231) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7642.245397494058, "metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "task": "social", "unit_id": "social/seed_0003"} +=== exited 2026-07-02T09:43:48.634923+00:00 returncode=0 === diff --git a/logs/social_seed_0004_attempt_01.log b/logs/social_seed_0004_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..1b5d6f41cd18efd92a1f2fadc55baaa6d6d7d1ed --- /dev/null +++ b/logs/social_seed_0004_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T07:36:16.525323+00:00 task=social seed=4 attempt=1 === +update=0001 collisions=1.70 chaser_return=-0.83 explorer_return=0.68 vision=(0.378,0.378) new_fields=(36.7,34.6) distance=5.05 loss=(-0.001,3.205) kl=-0.0000 +update=0500 collisions=6.38 chaser_return=8.84 explorer_return=42.36 vision=(0.460,0.460) new_fields=(60.2,65.0) distance=4.21 loss=(-0.006,5.577) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=12.53 chaser_return=14.09 explorer_return=34.83 vision=(0.582,0.582) new_fields=(53.2,62.9) distance=3.61 loss=(-0.007,6.011) kl=0.0005 +update=1500 collisions=12.03 chaser_return=12.55 explorer_return=33.24 vision=(0.725,0.725) new_fields=(49.0,62.2) distance=3.14 loss=(-0.007,5.755) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=14.45 chaser_return=15.05 explorer_return=28.31 vision=(0.835,0.835) new_fields=(48.7,59.5) distance=2.22 loss=(-0.006,5.787) kl=0.0008 +update=2500 collisions=13.50 chaser_return=14.67 explorer_return=30.70 vision=(0.835,0.835) new_fields=(52.0,60.9) distance=2.56 loss=(-0.006,5.782) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=21.48 chaser_return=21.84 explorer_return=23.56 vision=(0.914,0.914) new_fields=(44.4,59.7) distance=1.73 loss=(-0.003,6.184) kl=0.0003 +update=3500 collisions=16.05 chaser_return=16.70 explorer_return=27.25 vision=(0.880,0.880) new_fields=(49.1,60.4) distance=1.93 loss=(-0.006,5.322) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=16.80 chaser_return=17.41 explorer_return=24.06 vision=(0.840,0.840) new_fields=(48.5,58.0) distance=1.94 loss=(-0.005,5.411) kl=0.0008 +update=4500 collisions=17.80 chaser_return=18.23 explorer_return=25.06 vision=(0.918,0.918) new_fields=(46.4,60.2) distance=1.92 loss=(-0.006,5.698) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=21.48 chaser_return=22.33 explorer_return=19.74 vision=(0.930,0.930) new_fields=(47.9,58.2) distance=1.64 loss=(-0.005,5.707) kl=0.0011 +update=5500 collisions=18.27 chaser_return=18.97 explorer_return=24.34 vision=(0.936,0.936) new_fields=(48.4,59.5) distance=1.70 loss=(-0.008,5.519) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=17.05 chaser_return=18.25 explorer_return=29.60 vision=(0.894,0.894) new_fields=(50.0,61.9) distance=1.89 loss=(-0.006,4.938) kl=0.0008 +update=6500 collisions=16.48 chaser_return=17.04 explorer_return=26.06 vision=(0.933,0.933) new_fields=(48.3,60.5) distance=1.66 loss=(-0.006,5.170) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=16.55 chaser_return=17.50 explorer_return=27.26 vision=(0.923,0.923) new_fields=(49.6,61.3) distance=1.84 loss=(-0.006,5.630) kl=0.0012 +update=7500 collisions=22.00 chaser_return=23.08 explorer_return=20.34 vision=(0.933,0.933) new_fields=(48.1,59.0) distance=1.63 loss=(-0.005,5.723) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=17.62 chaser_return=18.70 explorer_return=28.41 vision=(0.923,0.923) new_fields=(50.4,63.1) distance=1.45 loss=(-0.004,5.002) kl=0.0004 +update=8500 collisions=20.15 chaser_return=21.70 explorer_return=24.11 vision=(0.941,0.941) new_fields=(51.0,61.0) distance=1.79 loss=(-0.005,4.883) kl=0.0016 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=19.48 chaser_return=20.04 explorer_return=23.96 vision=(0.930,0.930) new_fields=(48.3,61.5) distance=2.08 loss=(-0.005,5.764) kl=0.0013 +update=9500 collisions=18.62 chaser_return=20.44 explorer_return=30.91 vision=(0.945,0.945) new_fields=(53.0,64.1) distance=1.67 loss=(-0.004,4.688) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=14.10 chaser_return=14.21 explorer_return=34.23 vision=(0.926,0.926) new_fields=(46.1,63.9) distance=2.01 loss=(-0.003,5.754) kl=0.0003 +update=10500 collisions=16.65 chaser_return=17.77 explorer_return=24.40 vision=(0.947,0.947) new_fields=(51.1,59.2) distance=1.61 loss=(-0.005,5.539) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=20.60 chaser_return=22.02 explorer_return=26.36 vision=(0.933,0.933) new_fields=(50.0,61.6) distance=1.73 loss=(-0.003,5.189) kl=0.0009 +update=11500 collisions=16.68 chaser_return=17.77 explorer_return=35.30 vision=(0.954,0.954) new_fields=(50.3,67.7) distance=1.71 loss=(-0.005,4.762) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=16.88 chaser_return=17.46 explorer_return=28.49 vision=(0.935,0.935) new_fields=(47.8,62.2) distance=1.98 loss=(-0.005,5.271) kl=0.0008 +update=12500 collisions=18.55 chaser_return=19.29 explorer_return=24.05 vision=(0.943,0.943) new_fields=(48.9,60.7) distance=1.67 loss=(-0.003,5.154) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=16.02 chaser_return=17.47 explorer_return=30.34 vision=(0.947,0.947) new_fields=(52.1,64.1) distance=1.71 loss=(-0.004,5.049) kl=0.0000 +update=13500 collisions=16.60 chaser_return=18.08 explorer_return=33.99 vision=(0.947,0.947) new_fields=(52.9,66.9) distance=1.72 loss=(-0.003,4.745) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=18.45 chaser_return=19.59 explorer_return=27.59 vision=(0.952,0.952) new_fields=(50.1,60.9) distance=2.01 loss=(-0.005,5.542) kl=0.0007 +update=14500 collisions=20.52 chaser_return=21.76 explorer_return=19.69 vision=(0.938,0.938) new_fields=(50.0,57.7) distance=1.65 loss=(-0.003,5.719) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=18.10 chaser_return=19.14 explorer_return=23.23 vision=(0.951,0.951) new_fields=(48.9,58.3) distance=1.80 loss=(-0.004,5.447) kl=0.0004 +update=15500 collisions=9.18 chaser_return=10.16 explorer_return=22.06 vision=(0.944,0.944) new_fields=(52.6,53.2) distance=2.47 loss=(-0.002,6.393) kl=-0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=25.98 chaser_return=26.84 explorer_return=8.82 vision=(0.940,0.940) new_fields=(44.2,51.6) distance=1.72 loss=(-0.005,6.720) kl=0.0008 +update=16500 collisions=20.18 chaser_return=20.67 explorer_return=17.27 vision=(0.945,0.945) new_fields=(45.8,55.1) distance=1.45 loss=(-0.005,5.433) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=21.25 chaser_return=22.05 explorer_return=17.04 vision=(0.941,0.941) new_fields=(48.5,55.2) distance=1.96 loss=(-0.005,5.166) kl=0.0008 +update=17500 collisions=17.95 chaser_return=18.55 explorer_return=21.01 vision=(0.935,0.935) new_fields=(47.4,56.9) distance=1.55 loss=(-0.005,5.306) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=23.90 chaser_return=24.92 explorer_return=11.62 vision=(0.945,0.945) new_fields=(47.5,53.8) distance=1.61 loss=(-0.002,4.897) kl=0.0001 +update=18500 collisions=18.73 chaser_return=19.05 explorer_return=15.71 vision=(0.951,0.951) new_fields=(45.0,54.1) distance=1.60 loss=(-0.005,5.166) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=15.98 chaser_return=16.43 explorer_return=18.85 vision=(0.886,0.886) new_fields=(47.9,54.1) distance=2.02 loss=(-0.005,5.096) kl=0.0011 +update=19500 collisions=20.58 chaser_return=21.39 explorer_return=27.01 vision=(0.934,0.934) new_fields=(47.5,62.9) distance=1.57 loss=(-0.003,4.533) kl=0.0000 +update=20000 collisions=19.12 chaser_return=19.42 explorer_return=19.30 vision=(0.923,0.923) new_fields=(45.6,56.9) distance=1.83 loss=(-0.005,5.130) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7640.287686598953, "metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "task": "social", "unit_id": "social/seed_0004"} +=== exited 2026-07-02T09:43:48.659872+00:00 returncode=0 === diff --git a/logs/social_seed_0005_attempt_01.log b/logs/social_seed_0005_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..0f324d44551c3130d3e57671a59ced4efad5832b --- /dev/null +++ b/logs/social_seed_0005_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:18.366281+00:00 task=social seed=5 attempt=1 === +update=0001 collisions=1.30 chaser_return=-1.92 explorer_return=1.51 vision=(0.305,0.305) new_fields=(33.5,35.0) distance=5.48 loss=(-0.001,3.134) kl=-0.0000 +update=0500 collisions=13.03 chaser_return=14.07 explorer_return=31.12 vision=(0.923,0.923) new_fields=(51.4,61.0) distance=1.82 loss=(-0.003,5.539) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=15.38 chaser_return=16.13 explorer_return=29.09 vision=(0.922,0.922) new_fields=(49.2,61.4) distance=1.61 loss=(-0.006,5.701) kl=0.0003 +update=1500 collisions=13.53 chaser_return=13.99 explorer_return=32.34 vision=(0.912,0.912) new_fields=(49.2,62.6) distance=1.86 loss=(-0.005,5.412) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=15.95 chaser_return=17.07 explorer_return=32.62 vision=(0.934,0.934) new_fields=(51.7,64.7) distance=1.64 loss=(-0.005,5.396) kl=0.0003 +update=2500 collisions=15.82 chaser_return=16.84 explorer_return=30.44 vision=(0.862,0.862) new_fields=(50.9,62.5) distance=2.38 loss=(-0.005,5.692) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=15.38 chaser_return=16.06 explorer_return=32.61 vision=(0.898,0.898) new_fields=(50.3,65.2) distance=2.14 loss=(-0.005,5.627) kl=0.0009 +update=3500 collisions=13.75 chaser_return=14.53 explorer_return=34.17 vision=(0.894,0.894) new_fields=(50.6,64.2) distance=2.00 loss=(-0.006,5.376) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=16.38 chaser_return=17.21 explorer_return=29.11 vision=(0.927,0.927) new_fields=(49.6,62.7) distance=1.93 loss=(-0.006,5.346) kl=0.0010 +update=4500 collisions=16.95 chaser_return=17.44 explorer_return=30.81 vision=(0.936,0.936) new_fields=(47.6,63.5) distance=1.81 loss=(-0.006,5.516) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=18.20 chaser_return=19.41 explorer_return=29.93 vision=(0.951,0.951) new_fields=(50.6,63.7) distance=1.94 loss=(-0.003,5.560) kl=0.0009 +update=5500 collisions=20.88 chaser_return=21.36 explorer_return=25.74 vision=(0.924,0.924) new_fields=(46.5,62.2) distance=1.70 loss=(-0.004,5.574) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=17.48 chaser_return=18.21 explorer_return=29.09 vision=(0.934,0.934) new_fields=(48.8,62.5) distance=1.75 loss=(-0.004,5.393) kl=0.0003 +update=6500 collisions=18.25 chaser_return=19.27 explorer_return=27.54 vision=(0.933,0.933) new_fields=(50.6,62.3) distance=1.53 loss=(-0.004,5.698) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=23.20 chaser_return=24.22 explorer_return=15.24 vision=(0.959,0.959) new_fields=(47.4,56.0) distance=1.47 loss=(-0.002,5.525) kl=0.0004 +update=7500 collisions=17.20 chaser_return=18.11 explorer_return=26.11 vision=(0.945,0.945) new_fields=(49.0,60.7) distance=1.74 loss=(-0.006,5.427) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=18.75 chaser_return=19.70 explorer_return=28.11 vision=(0.936,0.936) new_fields=(50.2,63.5) distance=1.72 loss=(-0.004,5.190) kl=0.0005 +update=8500 collisions=21.68 chaser_return=23.25 explorer_return=25.19 vision=(0.955,0.955) new_fields=(53.2,64.1) distance=1.50 loss=(-0.002,5.458) kl=-0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=23.80 chaser_return=25.32 explorer_return=21.84 vision=(0.933,0.933) new_fields=(51.4,62.1) distance=1.66 loss=(-0.003,5.290) kl=0.0003 +update=9500 collisions=22.35 chaser_return=23.66 explorer_return=21.25 vision=(0.949,0.949) new_fields=(50.8,60.6) distance=1.57 loss=(-0.004,5.181) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=20.27 chaser_return=21.33 explorer_return=20.11 vision=(0.938,0.938) new_fields=(48.9,58.4) distance=1.55 loss=(-0.005,5.533) kl=0.0007 +update=10500 collisions=21.23 chaser_return=22.04 explorer_return=23.27 vision=(0.920,0.920) new_fields=(47.7,60.8) distance=1.86 loss=(-0.005,5.430) kl=0.0016 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=22.68 chaser_return=23.52 explorer_return=18.65 vision=(0.935,0.935) new_fields=(46.9,57.2) distance=1.66 loss=(-0.006,5.623) kl=0.0013 +update=11500 collisions=22.75 chaser_return=23.92 explorer_return=21.09 vision=(0.944,0.944) new_fields=(48.9,61.4) distance=1.77 loss=(-0.004,5.128) kl=0.0014 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=20.45 chaser_return=20.45 explorer_return=21.41 vision=(0.925,0.925) new_fields=(43.4,59.6) distance=1.70 loss=(-0.005,5.113) kl=0.0011 +update=12500 collisions=21.90 chaser_return=23.28 explorer_return=24.59 vision=(0.938,0.938) new_fields=(50.8,63.5) distance=1.39 loss=(-0.004,5.261) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=20.43 chaser_return=21.03 explorer_return=19.77 vision=(0.931,0.931) new_fields=(46.5,58.6) distance=1.57 loss=(-0.004,5.448) kl=0.0010 +update=13500 collisions=19.02 chaser_return=20.70 explorer_return=30.41 vision=(0.954,0.954) new_fields=(53.2,65.9) distance=1.36 loss=(-0.004,5.111) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=22.40 chaser_return=22.98 explorer_return=17.06 vision=(0.948,0.948) new_fields=(46.5,58.0) distance=1.60 loss=(-0.006,5.585) kl=0.0009 +update=14500 collisions=19.43 chaser_return=20.25 explorer_return=25.08 vision=(0.940,0.940) new_fields=(48.6,61.8) distance=1.63 loss=(-0.003,5.608) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=18.88 chaser_return=19.23 explorer_return=22.76 vision=(0.926,0.926) new_fields=(46.3,59.2) distance=1.96 loss=(-0.004,5.758) kl=0.0005 +update=15500 collisions=22.75 chaser_return=23.76 explorer_return=22.18 vision=(0.934,0.934) new_fields=(47.8,59.2) distance=1.78 loss=(-0.005,5.995) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=20.80 chaser_return=22.15 explorer_return=23.83 vision=(0.963,0.963) new_fields=(50.9,61.7) distance=1.65 loss=(-0.003,5.587) kl=0.0006 +update=16500 collisions=21.12 chaser_return=22.20 explorer_return=24.19 vision=(0.952,0.952) new_fields=(48.8,61.3) distance=1.58 loss=(-0.004,5.014) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=19.77 chaser_return=21.96 explorer_return=31.31 vision=(0.951,0.951) new_fields=(55.0,66.8) distance=1.62 loss=(-0.003,4.686) kl=0.0004 +update=17500 collisions=18.55 chaser_return=19.74 explorer_return=24.80 vision=(0.950,0.950) new_fields=(49.8,61.0) distance=1.75 loss=(-0.005,4.887) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=19.40 chaser_return=20.70 explorer_return=22.91 vision=(0.959,0.959) new_fields=(50.5,59.9) distance=1.65 loss=(-0.003,5.135) kl=0.0003 +update=18500 collisions=22.45 chaser_return=23.47 explorer_return=24.76 vision=(0.948,0.948) new_fields=(48.6,62.8) distance=1.47 loss=(-0.002,5.429) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=20.02 chaser_return=21.27 explorer_return=23.01 vision=(0.936,0.936) new_fields=(50.7,58.5) distance=1.65 loss=(-0.004,5.677) kl=0.0002 +update=19500 collisions=16.45 chaser_return=17.51 explorer_return=29.38 vision=(0.922,0.922) new_fields=(50.2,62.4) distance=1.90 loss=(-0.005,4.853) kl=0.0006 +update=20000 collisions=18.58 chaser_return=20.12 explorer_return=20.77 vision=(0.953,0.953) new_fields=(51.5,58.6) distance=1.58 loss=(-0.003,4.460) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7556.884778791922, "metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "task": "social", "unit_id": "social/seed_0005"} +=== exited 2026-07-02T11:49:20.760552+00:00 returncode=0 === diff --git a/logs/social_seed_0006_attempt_01.log b/logs/social_seed_0006_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..70117dee4d0473a59ceec5a6c8eafa91ce6f4d16 --- /dev/null +++ b/logs/social_seed_0006_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:38.590438+00:00 task=social seed=6 attempt=1 === +update=0001 collisions=2.45 chaser_return=-0.59 explorer_return=-1.69 vision=(0.423,0.423) new_fields=(34.0,33.5) distance=4.61 loss=(-0.001,3.144) kl=-0.0001 +update=0500 collisions=10.60 chaser_return=11.59 explorer_return=32.08 vision=(0.866,0.866) new_fields=(51.8,60.5) distance=2.10 loss=(-0.006,5.398) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=10.78 chaser_return=11.69 explorer_return=34.73 vision=(0.892,0.892) new_fields=(51.7,62.7) distance=2.16 loss=(-0.007,5.382) kl=0.0011 +update=1500 collisions=16.58 chaser_return=18.04 explorer_return=25.56 vision=(0.913,0.913) new_fields=(51.8,59.0) distance=2.27 loss=(-0.005,5.727) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=14.12 chaser_return=15.67 explorer_return=35.53 vision=(0.914,0.914) new_fields=(53.6,64.7) distance=1.87 loss=(-0.007,5.695) kl=0.0009 +update=2500 collisions=15.82 chaser_return=16.51 explorer_return=28.52 vision=(0.895,0.895) new_fields=(48.5,61.3) distance=1.95 loss=(-0.006,5.511) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=15.85 chaser_return=16.76 explorer_return=30.05 vision=(0.934,0.934) new_fields=(50.0,62.4) distance=1.69 loss=(-0.005,5.600) kl=0.0003 +update=3500 collisions=20.73 chaser_return=21.29 explorer_return=23.00 vision=(0.905,0.905) new_fields=(45.4,58.6) distance=1.98 loss=(-0.004,5.874) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=17.58 chaser_return=17.99 explorer_return=24.01 vision=(0.906,0.906) new_fields=(46.7,59.8) distance=2.17 loss=(-0.006,5.487) kl=0.0007 +update=4500 collisions=17.98 chaser_return=18.74 explorer_return=26.29 vision=(0.922,0.922) new_fields=(49.6,61.5) distance=1.86 loss=(-0.006,5.626) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=16.38 chaser_return=17.31 explorer_return=32.90 vision=(0.915,0.915) new_fields=(50.2,65.5) distance=1.74 loss=(-0.006,5.136) kl=0.0010 +update=5500 collisions=17.20 chaser_return=18.22 explorer_return=29.71 vision=(0.930,0.930) new_fields=(50.0,63.4) distance=1.70 loss=(-0.005,5.252) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=19.43 chaser_return=20.94 explorer_return=24.40 vision=(0.941,0.941) new_fields=(51.2,61.4) distance=1.44 loss=(-0.006,5.647) kl=0.0006 +update=6500 collisions=19.00 chaser_return=20.03 explorer_return=26.34 vision=(0.946,0.946) new_fields=(49.3,62.2) distance=1.66 loss=(-0.008,5.644) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=18.08 chaser_return=19.50 explorer_return=27.93 vision=(0.935,0.935) new_fields=(51.8,63.1) distance=1.62 loss=(-0.005,5.577) kl=0.0023 +update=7500 collisions=16.90 chaser_return=17.54 explorer_return=27.73 vision=(0.872,0.872) new_fields=(48.5,61.9) distance=1.98 loss=(-0.006,5.703) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=17.93 chaser_return=18.60 explorer_return=23.91 vision=(0.919,0.919) new_fields=(47.8,59.9) distance=1.55 loss=(-0.004,5.587) kl=0.0002 +update=8500 collisions=15.35 chaser_return=16.16 explorer_return=28.50 vision=(0.913,0.913) new_fields=(49.8,61.7) distance=2.01 loss=(-0.006,5.374) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=22.93 chaser_return=23.25 explorer_return=19.39 vision=(0.958,0.958) new_fields=(43.9,59.0) distance=1.75 loss=(-0.005,5.772) kl=0.0009 +update=9500 collisions=18.52 chaser_return=19.30 explorer_return=23.54 vision=(0.916,0.916) new_fields=(49.0,59.7) distance=2.08 loss=(-0.006,5.346) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=17.85 chaser_return=19.39 explorer_return=26.76 vision=(0.943,0.943) new_fields=(53.2,61.2) distance=1.79 loss=(-0.005,5.027) kl=0.0010 +update=10500 collisions=17.80 chaser_return=18.59 explorer_return=20.38 vision=(0.945,0.945) new_fields=(48.1,57.2) distance=1.84 loss=(-0.006,5.263) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=19.60 chaser_return=21.08 explorer_return=27.39 vision=(0.935,0.935) new_fields=(51.7,63.4) distance=1.69 loss=(-0.003,4.567) kl=0.0003 +update=11500 collisions=19.65 chaser_return=20.52 explorer_return=26.12 vision=(0.933,0.933) new_fields=(48.6,62.0) distance=1.70 loss=(-0.005,5.235) kl=0.0016 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=26.08 chaser_return=27.74 explorer_return=18.04 vision=(0.958,0.958) new_fields=(50.5,59.3) distance=1.64 loss=(-0.005,5.639) kl=0.0009 +update=12500 collisions=20.52 chaser_return=22.68 explorer_return=34.31 vision=(0.956,0.956) new_fields=(55.2,68.9) distance=1.74 loss=(-0.004,5.144) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=27.95 chaser_return=29.96 explorer_return=18.00 vision=(0.955,0.955) new_fields=(52.7,61.4) distance=1.54 loss=(-0.005,5.030) kl=0.0010 +update=13500 collisions=18.88 chaser_return=19.42 explorer_return=19.46 vision=(0.938,0.938) new_fields=(46.8,57.2) distance=1.64 loss=(-0.005,5.378) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=19.75 chaser_return=21.11 explorer_return=25.74 vision=(0.955,0.955) new_fields=(51.2,61.6) distance=1.57 loss=(-0.003,5.500) kl=-0.0000 +update=14500 collisions=19.60 chaser_return=21.12 explorer_return=23.38 vision=(0.951,0.951) new_fields=(51.9,60.0) distance=1.67 loss=(-0.004,5.271) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=17.85 chaser_return=18.62 explorer_return=25.52 vision=(0.957,0.957) new_fields=(47.9,61.5) distance=2.02 loss=(-0.003,4.485) kl=0.0002 +update=15500 collisions=18.95 chaser_return=20.85 explorer_return=26.70 vision=(0.951,0.951) new_fields=(52.9,62.6) distance=1.57 loss=(-0.003,4.751) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=24.48 chaser_return=26.91 explorer_return=25.29 vision=(0.939,0.939) new_fields=(53.7,63.5) distance=1.58 loss=(-0.004,4.981) kl=0.0008 +update=16500 collisions=32.12 chaser_return=33.47 explorer_return=7.62 vision=(0.945,0.945) new_fields=(44.3,53.1) distance=1.35 loss=(-0.003,6.302) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=18.62 chaser_return=19.92 explorer_return=23.94 vision=(0.933,0.933) new_fields=(51.0,60.4) distance=2.08 loss=(-0.005,5.160) kl=0.0005 +update=17500 collisions=24.20 chaser_return=25.67 explorer_return=18.79 vision=(0.941,0.941) new_fields=(48.7,59.0) distance=1.74 loss=(-0.004,5.314) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=22.60 chaser_return=23.50 explorer_return=17.56 vision=(0.955,0.955) new_fields=(47.3,57.0) distance=1.60 loss=(-0.005,5.722) kl=0.0001 +update=18500 collisions=18.80 chaser_return=19.51 explorer_return=21.90 vision=(0.952,0.952) new_fields=(47.6,57.5) distance=1.68 loss=(-0.004,5.387) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=17.95 chaser_return=18.95 explorer_return=24.99 vision=(0.941,0.941) new_fields=(50.0,59.9) distance=1.50 loss=(-0.005,4.928) kl=0.0005 +update=19500 collisions=18.70 chaser_return=19.60 explorer_return=21.35 vision=(0.947,0.947) new_fields=(49.1,58.4) distance=1.58 loss=(-0.004,5.470) kl=0.0009 +update=20000 collisions=19.80 chaser_return=21.02 explorer_return=22.49 vision=(0.925,0.925) new_fields=(49.9,59.4) distance=1.64 loss=(-0.004,4.686) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7558.831456203014, "metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "task": "social", "unit_id": "social/seed_0006"} +=== exited 2026-07-02T11:49:40.902414+00:00 returncode=0 === diff --git a/logs/social_seed_0007_attempt_01.log b/logs/social_seed_0007_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..8067c998021628e0955c35adaa0704f4b6b2718e --- /dev/null +++ b/logs/social_seed_0007_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:48.618172+00:00 task=social seed=7 attempt=1 === +update=0001 collisions=1.38 chaser_return=-1.75 explorer_return=0.99 vision=(0.339,0.339) new_fields=(34.1,34.7) distance=5.21 loss=(-0.001,3.165) kl=0.0000 +update=0500 collisions=8.05 chaser_return=10.61 explorer_return=38.71 vision=(0.504,0.504) new_fields=(60.2,63.5) distance=3.50 loss=(-0.006,5.447) kl=0.0012 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=11.03 chaser_return=13.36 explorer_return=34.98 vision=(0.539,0.539) new_fields=(58.0,62.2) distance=3.88 loss=(-0.006,6.057) kl=0.0001 +update=1500 collisions=12.70 chaser_return=14.61 explorer_return=33.54 vision=(0.561,0.561) new_fields=(55.0,62.1) distance=3.62 loss=(-0.004,5.964) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=13.03 chaser_return=13.92 explorer_return=31.73 vision=(0.689,0.689) new_fields=(50.5,61.6) distance=2.93 loss=(-0.007,5.827) kl=0.0001 +update=2500 collisions=13.05 chaser_return=13.67 explorer_return=29.69 vision=(0.764,0.764) new_fields=(49.1,60.1) distance=2.39 loss=(-0.007,5.747) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=16.40 chaser_return=17.32 explorer_return=29.81 vision=(0.874,0.874) new_fields=(50.9,62.9) distance=1.84 loss=(-0.005,5.488) kl=-0.0001 +update=3500 collisions=16.30 chaser_return=17.11 explorer_return=29.11 vision=(0.945,0.945) new_fields=(49.8,61.5) distance=1.91 loss=(-0.005,5.188) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=17.27 chaser_return=17.98 explorer_return=25.36 vision=(0.911,0.911) new_fields=(49.0,60.2) distance=2.09 loss=(-0.005,5.741) kl=-0.0002 +update=4500 collisions=17.38 chaser_return=18.45 explorer_return=29.66 vision=(0.940,0.940) new_fields=(49.7,63.3) distance=1.76 loss=(-0.006,5.378) kl=0.0013 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=17.55 chaser_return=17.68 explorer_return=22.56 vision=(0.928,0.928) new_fields=(45.8,58.2) distance=1.77 loss=(-0.006,5.665) kl=0.0003 +update=5500 collisions=15.15 chaser_return=16.34 explorer_return=32.12 vision=(0.945,0.945) new_fields=(52.0,64.8) distance=1.70 loss=(-0.004,5.231) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=17.43 chaser_return=18.39 explorer_return=25.70 vision=(0.920,0.920) new_fields=(49.5,59.9) distance=1.98 loss=(-0.006,5.880) kl=0.0006 +update=6500 collisions=18.08 chaser_return=18.49 explorer_return=20.99 vision=(0.941,0.941) new_fields=(47.1,57.9) distance=1.81 loss=(-0.007,5.635) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=19.62 chaser_return=21.00 explorer_return=24.49 vision=(0.954,0.954) new_fields=(50.6,60.3) distance=1.67 loss=(-0.003,5.471) kl=0.0006 +update=7500 collisions=18.20 chaser_return=19.45 explorer_return=34.46 vision=(0.946,0.946) new_fields=(51.3,66.5) distance=1.81 loss=(-0.004,5.055) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=19.08 chaser_return=19.99 explorer_return=26.34 vision=(0.922,0.922) new_fields=(48.8,62.2) distance=1.77 loss=(-0.004,5.324) kl=0.0007 +update=8500 collisions=18.35 chaser_return=18.99 explorer_return=25.39 vision=(0.953,0.953) new_fields=(48.6,61.3) distance=1.76 loss=(-0.006,5.502) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=15.73 chaser_return=16.87 explorer_return=28.80 vision=(0.930,0.930) new_fields=(51.5,61.9) distance=1.64 loss=(-0.006,5.099) kl=0.0006 +update=9500 collisions=22.27 chaser_return=22.89 explorer_return=18.89 vision=(0.914,0.914) new_fields=(46.2,57.4) distance=1.82 loss=(-0.004,5.842) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=17.08 chaser_return=18.38 explorer_return=32.36 vision=(0.955,0.955) new_fields=(52.2,64.9) distance=1.55 loss=(-0.002,5.039) kl=-0.0003 +update=10500 collisions=20.27 chaser_return=20.30 explorer_return=20.38 vision=(0.951,0.951) new_fields=(43.8,59.0) distance=1.77 loss=(-0.005,5.443) kl=0.0019 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=24.60 chaser_return=25.43 explorer_return=19.15 vision=(0.947,0.947) new_fields=(46.5,59.5) distance=1.52 loss=(-0.006,5.679) kl=0.0012 +update=11500 collisions=20.80 chaser_return=21.65 explorer_return=19.48 vision=(0.934,0.934) new_fields=(48.4,57.9) distance=1.78 loss=(-0.004,5.281) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=23.35 chaser_return=24.62 explorer_return=20.00 vision=(0.910,0.910) new_fields=(50.2,60.8) distance=1.63 loss=(-0.003,5.122) kl=0.0003 +update=12500 collisions=20.23 chaser_return=21.25 explorer_return=23.21 vision=(0.943,0.943) new_fields=(50.1,61.1) distance=1.90 loss=(-0.003,4.965) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=19.23 chaser_return=19.65 explorer_return=10.25 vision=(0.953,0.953) new_fields=(46.8,50.4) distance=1.93 loss=(-0.003,5.366) kl=-0.0003 +update=13500 collisions=20.02 chaser_return=20.71 explorer_return=23.76 vision=(0.942,0.942) new_fields=(47.7,60.1) distance=1.68 loss=(-0.005,5.469) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=19.80 chaser_return=20.08 explorer_return=24.21 vision=(0.942,0.942) new_fields=(44.8,60.4) distance=2.14 loss=(-0.005,5.046) kl=0.0013 +update=14500 collisions=16.70 chaser_return=16.76 explorer_return=28.01 vision=(0.911,0.911) new_fields=(46.1,62.4) distance=1.93 loss=(-0.004,5.335) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=20.40 chaser_return=21.94 explorer_return=24.59 vision=(0.941,0.941) new_fields=(51.5,61.7) distance=1.65 loss=(-0.004,5.087) kl=0.0004 +update=15500 collisions=23.65 chaser_return=24.59 explorer_return=17.49 vision=(0.930,0.930) new_fields=(47.6,58.2) distance=1.64 loss=(-0.004,4.988) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=20.38 chaser_return=21.50 explorer_return=22.20 vision=(0.936,0.936) new_fields=(49.5,60.7) distance=1.61 loss=(-0.005,4.797) kl=0.0009 +update=16500 collisions=18.48 chaser_return=19.23 explorer_return=24.99 vision=(0.919,0.919) new_fields=(48.9,61.4) distance=1.68 loss=(-0.006,5.468) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=19.12 chaser_return=20.75 explorer_return=27.89 vision=(0.926,0.926) new_fields=(54.1,61.5) distance=1.80 loss=(-0.002,5.051) kl=0.0004 +update=17500 collisions=19.75 chaser_return=20.71 explorer_return=20.98 vision=(0.935,0.935) new_fields=(48.5,58.0) distance=1.58 loss=(-0.006,5.685) kl=0.0016 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=18.50 chaser_return=19.34 explorer_return=21.08 vision=(0.943,0.943) new_fields=(48.0,58.2) distance=1.59 loss=(-0.003,4.998) kl=-0.0002 +update=18500 collisions=21.93 chaser_return=22.19 explorer_return=9.43 vision=(0.932,0.932) new_fields=(42.9,50.0) distance=1.68 loss=(-0.004,5.543) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=18.02 chaser_return=18.05 explorer_return=18.01 vision=(0.945,0.945) new_fields=(45.5,55.5) distance=1.52 loss=(-0.003,5.345) kl=0.0006 +update=19500 collisions=18.75 chaser_return=19.64 explorer_return=21.66 vision=(0.925,0.925) new_fields=(49.4,58.0) distance=1.95 loss=(-0.005,5.504) kl=0.0005 +update=20000 collisions=23.20 chaser_return=24.29 explorer_return=17.15 vision=(0.947,0.947) new_fields=(46.9,55.9) distance=1.69 loss=(-0.003,4.716) kl=0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7587.021667060093, "metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "task": "social", "unit_id": "social/seed_0007"} +=== exited 2026-07-02T11:50:20.928628+00:00 returncode=0 === diff --git a/logs/social_seed_0008_attempt_01.log b/logs/social_seed_0008_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..5d56ed15d69b5f9790440d07750ff15d1a424b14 --- /dev/null +++ b/logs/social_seed_0008_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:58.682914+00:00 task=social seed=8 attempt=1 === +update=0001 collisions=1.20 chaser_return=-2.02 explorer_return=-1.06 vision=(0.313,0.313) new_fields=(33.5,33.2) distance=5.32 loss=(-0.001,3.142) kl=0.0000 +update=0500 collisions=6.83 chaser_return=9.10 explorer_return=39.33 vision=(0.550,0.550) new_fields=(59.3,63.1) distance=4.29 loss=(-0.006,5.485) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=11.62 chaser_return=12.99 explorer_return=35.20 vision=(0.729,0.729) new_fields=(53.3,62.7) distance=2.64 loss=(-0.004,5.647) kl=0.0006 +update=1500 collisions=12.00 chaser_return=12.21 explorer_return=29.58 vision=(0.801,0.801) new_fields=(47.3,59.5) distance=1.95 loss=(-0.007,5.451) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=10.90 chaser_return=11.75 explorer_return=35.08 vision=(0.868,0.868) new_fields=(51.8,63.0) distance=2.37 loss=(-0.006,5.021) kl=0.0001 +update=2500 collisions=12.75 chaser_return=13.52 explorer_return=31.94 vision=(0.903,0.903) new_fields=(50.3,62.0) distance=1.81 loss=(-0.007,5.398) kl=0.0012 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=13.90 chaser_return=14.72 explorer_return=30.65 vision=(0.892,0.892) new_fields=(49.7,61.9) distance=2.06 loss=(-0.006,5.417) kl=0.0010 +update=3500 collisions=14.68 chaser_return=15.10 explorer_return=31.76 vision=(0.869,0.869) new_fields=(47.8,63.0) distance=2.22 loss=(-0.006,5.582) kl=0.0018 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=15.60 chaser_return=16.34 explorer_return=29.99 vision=(0.898,0.898) new_fields=(49.5,62.4) distance=2.12 loss=(-0.005,5.858) kl=0.0008 +update=4500 collisions=15.15 chaser_return=15.73 explorer_return=28.71 vision=(0.889,0.889) new_fields=(47.9,61.0) distance=1.94 loss=(-0.004,5.343) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=17.00 chaser_return=17.99 explorer_return=28.76 vision=(0.889,0.889) new_fields=(49.3,61.7) distance=1.86 loss=(-0.005,5.814) kl=0.0008 +update=5500 collisions=19.68 chaser_return=21.23 explorer_return=32.04 vision=(0.915,0.915) new_fields=(50.9,65.5) distance=1.87 loss=(-0.005,5.346) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=18.52 chaser_return=19.54 explorer_return=22.49 vision=(0.933,0.933) new_fields=(48.6,58.5) distance=1.71 loss=(-0.006,5.770) kl=0.0014 +update=6500 collisions=19.00 chaser_return=19.54 explorer_return=23.56 vision=(0.918,0.918) new_fields=(48.0,60.2) distance=1.92 loss=(-0.005,5.712) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=18.23 chaser_return=19.06 explorer_return=23.27 vision=(0.894,0.894) new_fields=(48.4,59.2) distance=1.81 loss=(-0.005,5.607) kl=0.0015 +update=7500 collisions=16.70 chaser_return=17.57 explorer_return=30.11 vision=(0.930,0.930) new_fields=(49.8,62.9) distance=1.84 loss=(-0.006,5.406) kl=0.0011 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=15.40 chaser_return=16.61 explorer_return=31.62 vision=(0.938,0.938) new_fields=(51.7,63.1) distance=1.77 loss=(-0.006,5.285) kl=0.0009 +update=8500 collisions=15.23 chaser_return=15.96 explorer_return=24.44 vision=(0.909,0.909) new_fields=(49.7,58.6) distance=1.98 loss=(-0.005,5.613) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=20.58 chaser_return=21.17 explorer_return=20.41 vision=(0.947,0.947) new_fields=(48.1,59.1) distance=1.81 loss=(-0.006,5.469) kl=0.0005 +update=9500 collisions=22.60 chaser_return=23.70 explorer_return=18.12 vision=(0.939,0.939) new_fields=(49.2,58.2) distance=1.60 loss=(-0.005,5.487) kl=0.0021 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=20.18 chaser_return=20.66 explorer_return=20.43 vision=(0.932,0.932) new_fields=(46.4,58.2) distance=1.64 loss=(-0.004,5.325) kl=0.0005 +update=10500 collisions=20.40 chaser_return=21.83 explorer_return=25.45 vision=(0.945,0.945) new_fields=(51.4,61.8) distance=1.43 loss=(-0.003,5.363) kl=-0.0001 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=18.85 chaser_return=19.87 explorer_return=24.01 vision=(0.936,0.936) new_fields=(49.7,61.0) distance=1.72 loss=(-0.005,5.403) kl=0.0013 +update=11500 collisions=16.52 chaser_return=17.88 explorer_return=22.02 vision=(0.944,0.944) new_fields=(51.5,58.3) distance=1.72 loss=(-0.005,5.244) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=19.05 chaser_return=20.25 explorer_return=22.56 vision=(0.951,0.951) new_fields=(50.1,58.5) distance=1.99 loss=(-0.005,5.343) kl=0.0011 +update=12500 collisions=18.18 chaser_return=18.68 explorer_return=26.49 vision=(0.892,0.892) new_fields=(47.2,62.0) distance=1.89 loss=(-0.006,5.339) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=17.65 chaser_return=19.17 explorer_return=25.33 vision=(0.938,0.938) new_fields=(52.0,60.5) distance=1.71 loss=(-0.004,5.004) kl=0.0008 +update=13500 collisions=18.50 chaser_return=18.90 explorer_return=17.77 vision=(0.944,0.944) new_fields=(47.2,55.9) distance=1.94 loss=(-0.003,5.405) kl=0.0008 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=17.95 chaser_return=19.27 explorer_return=19.81 vision=(0.926,0.926) new_fields=(51.1,57.3) distance=1.83 loss=(-0.003,5.690) kl=0.0006 +update=14500 collisions=17.95 chaser_return=18.61 explorer_return=23.60 vision=(0.912,0.912) new_fields=(47.5,59.5) distance=1.85 loss=(-0.002,4.620) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=27.30 chaser_return=28.75 explorer_return=18.36 vision=(0.904,0.904) new_fields=(47.0,57.9) distance=1.60 loss=(0.002,5.836) kl=0.0139 +update=15500 collisions=17.55 chaser_return=18.63 explorer_return=27.77 vision=(0.902,0.902) new_fields=(50.2,62.0) distance=1.71 loss=(-0.003,4.693) kl=-0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=18.27 chaser_return=19.75 explorer_return=24.79 vision=(0.914,0.914) new_fields=(51.7,59.4) distance=1.78 loss=(-0.002,4.726) kl=0.0004 +update=16500 collisions=18.73 chaser_return=20.18 explorer_return=19.39 vision=(0.906,0.906) new_fields=(51.2,55.9) distance=1.95 loss=(-0.004,5.000) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=15.98 chaser_return=17.19 explorer_return=28.15 vision=(0.916,0.916) new_fields=(51.8,61.7) distance=1.78 loss=(-0.004,5.103) kl=0.0010 +update=17500 collisions=20.90 chaser_return=22.00 explorer_return=18.19 vision=(0.922,0.922) new_fields=(48.6,57.2) distance=1.83 loss=(-0.004,5.165) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=19.90 chaser_return=21.10 explorer_return=21.99 vision=(0.878,0.878) new_fields=(49.7,59.7) distance=1.80 loss=(-0.005,5.162) kl=0.0011 +update=18500 collisions=22.43 chaser_return=23.93 explorer_return=18.33 vision=(0.921,0.921) new_fields=(50.5,57.9) distance=1.66 loss=(-0.004,5.221) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=22.15 chaser_return=22.85 explorer_return=15.20 vision=(0.855,0.855) new_fields=(45.8,54.5) distance=2.01 loss=(-0.004,5.898) kl=0.0006 +update=19500 collisions=17.43 chaser_return=18.17 explorer_return=26.64 vision=(0.907,0.907) new_fields=(48.9,61.5) distance=2.11 loss=(-0.004,4.408) kl=0.0004 +update=20000 collisions=19.68 chaser_return=20.85 explorer_return=23.00 vision=(0.921,0.921) new_fields=(50.1,60.5) distance=1.83 loss=(-0.005,4.698) kl=0.0023 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7584.432566124015, "metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "task": "social", "unit_id": "social/seed_0008"} +=== exited 2026-07-02T11:50:30.952355+00:00 returncode=0 === diff --git a/logs/social_seed_0009_attempt_01.log b/logs/social_seed_0009_attempt_01.log new file mode 100644 index 0000000000000000000000000000000000000000..ec8a8ec8bf562fc5b6387a53399e7d91b6e6c918 --- /dev/null +++ b/logs/social_seed_0009_attempt_01.log @@ -0,0 +1,65 @@ + +=== started 2026-07-02T09:43:58.684577+00:00 task=social seed=9 attempt=1 === +update=0001 collisions=1.38 chaser_return=-1.62 explorer_return=2.15 vision=(0.315,0.315) new_fields=(34.5,35.5) distance=4.80 loss=(-0.001,3.074) kl=0.0001 +update=0500 collisions=6.12 chaser_return=8.67 explorer_return=40.76 vision=(0.493,0.493) new_fields=(60.8,63.8) distance=4.03 loss=(-0.006,5.433) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_001000.safetensors +update=1000 collisions=11.48 chaser_return=12.86 explorer_return=34.00 vision=(0.626,0.626) new_fields=(53.2,62.2) distance=3.38 loss=(-0.007,5.897) kl=0.0007 +update=1500 collisions=11.40 chaser_return=12.35 explorer_return=35.54 vision=(0.765,0.765) new_fields=(51.5,63.3) distance=2.84 loss=(-0.008,5.701) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_002000.safetensors +update=2000 collisions=10.35 chaser_return=11.32 explorer_return=33.70 vision=(0.746,0.746) new_fields=(51.9,61.4) distance=2.18 loss=(-0.005,5.547) kl=0.0002 +update=2500 collisions=12.48 chaser_return=13.42 explorer_return=31.74 vision=(0.672,0.672) new_fields=(50.6,60.9) distance=2.45 loss=(-0.004,5.377) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_003000.safetensors +update=3000 collisions=13.95 chaser_return=13.94 explorer_return=31.23 vision=(0.810,0.810) new_fields=(46.0,61.7) distance=2.36 loss=(-0.008,5.507) kl=0.0004 +update=3500 collisions=19.62 chaser_return=20.07 explorer_return=24.30 vision=(0.900,0.900) new_fields=(46.2,59.9) distance=1.85 loss=(-0.005,5.836) kl=0.0004 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_004000.safetensors +update=4000 collisions=18.08 chaser_return=19.12 explorer_return=28.11 vision=(0.922,0.922) new_fields=(50.0,62.3) distance=1.74 loss=(-0.005,5.758) kl=0.0006 +update=4500 collisions=19.73 chaser_return=20.26 explorer_return=24.51 vision=(0.879,0.879) new_fields=(46.0,59.5) distance=2.18 loss=(-0.004,6.040) kl=0.0012 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_005000.safetensors +update=5000 collisions=18.12 chaser_return=18.81 explorer_return=25.65 vision=(0.912,0.912) new_fields=(48.0,60.9) distance=1.92 loss=(-0.005,5.578) kl=0.0008 +update=5500 collisions=16.12 chaser_return=17.06 explorer_return=32.31 vision=(0.903,0.903) new_fields=(50.2,65.0) distance=1.90 loss=(-0.005,5.384) kl=0.0006 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_006000.safetensors +update=6000 collisions=16.58 chaser_return=17.38 explorer_return=31.04 vision=(0.928,0.928) new_fields=(49.6,63.4) distance=1.78 loss=(-0.006,5.411) kl=0.0009 +update=6500 collisions=15.32 chaser_return=16.75 explorer_return=31.74 vision=(0.926,0.926) new_fields=(53.5,63.4) distance=1.94 loss=(-0.007,5.397) kl=0.0015 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_007000.safetensors +update=7000 collisions=15.05 chaser_return=16.26 explorer_return=26.51 vision=(0.930,0.930) new_fields=(52.4,59.9) distance=1.82 loss=(-0.006,5.528) kl=0.0013 +update=7500 collisions=19.75 chaser_return=21.01 explorer_return=26.34 vision=(0.923,0.923) new_fields=(50.7,61.7) distance=1.75 loss=(-0.003,5.391) kl=0.0002 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_008000.safetensors +update=8000 collisions=18.25 chaser_return=18.58 explorer_return=21.50 vision=(0.936,0.936) new_fields=(46.8,58.3) distance=1.65 loss=(-0.006,5.550) kl=0.0009 +update=8500 collisions=18.18 chaser_return=19.55 explorer_return=23.19 vision=(0.937,0.937) new_fields=(50.9,58.9) distance=2.01 loss=(-0.004,5.536) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_009000.safetensors +update=9000 collisions=19.70 chaser_return=20.76 explorer_return=23.81 vision=(0.896,0.896) new_fields=(49.4,59.3) distance=2.05 loss=(-0.006,6.206) kl=0.0010 +update=9500 collisions=20.40 chaser_return=19.84 explorer_return=4.94 vision=(0.936,0.936) new_fields=(39.7,46.7) distance=1.39 loss=(-0.003,5.978) kl=0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_010000.safetensors +update=10000 collisions=3.60 chaser_return=-0.96 explorer_return=52.08 vision=(0.477,0.477) new_fields=(26.4,70.8) distance=5.93 loss=(-0.002,6.723) kl=-0.0005 +update=10500 collisions=16.50 chaser_return=17.14 explorer_return=27.18 vision=(0.912,0.912) new_fields=(47.9,60.6) distance=1.75 loss=(-0.003,5.247) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_011000.safetensors +update=11000 collisions=19.65 chaser_return=20.23 explorer_return=17.12 vision=(0.940,0.940) new_fields=(47.0,55.2) distance=1.72 loss=(-0.006,5.099) kl=0.0014 +update=11500 collisions=18.83 chaser_return=20.05 explorer_return=30.06 vision=(0.955,0.955) new_fields=(51.0,63.2) distance=1.88 loss=(-0.004,4.981) kl=0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_012000.safetensors +update=12000 collisions=18.08 chaser_return=18.65 explorer_return=22.34 vision=(0.907,0.907) new_fields=(47.4,58.2) distance=1.80 loss=(-0.003,4.145) kl=0.0004 +update=12500 collisions=19.98 chaser_return=20.64 explorer_return=24.54 vision=(0.942,0.942) new_fields=(47.7,61.1) distance=1.73 loss=(-0.004,4.642) kl=0.0010 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_013000.safetensors +update=13000 collisions=19.62 chaser_return=20.90 explorer_return=21.83 vision=(0.910,0.910) new_fields=(49.8,58.3) distance=1.62 loss=(-0.004,5.183) kl=-0.0001 +update=13500 collisions=26.38 chaser_return=27.62 explorer_return=16.20 vision=(0.914,0.914) new_fields=(47.5,57.1) distance=1.58 loss=(-0.004,5.743) kl=-0.0000 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_014000.safetensors +update=14000 collisions=18.58 chaser_return=20.01 explorer_return=24.71 vision=(0.923,0.923) new_fields=(51.2,59.9) distance=1.70 loss=(-0.005,4.892) kl=0.0010 +update=14500 collisions=21.05 chaser_return=22.18 explorer_return=27.41 vision=(0.911,0.911) new_fields=(48.8,63.1) distance=2.06 loss=(-0.004,4.936) kl=0.0005 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_015000.safetensors +update=15000 collisions=15.12 chaser_return=15.53 explorer_return=25.88 vision=(0.941,0.941) new_fields=(47.5,59.5) distance=1.76 loss=(-0.005,5.138) kl=0.0009 +update=15500 collisions=20.98 chaser_return=21.89 explorer_return=22.80 vision=(0.932,0.932) new_fields=(47.0,61.3) distance=1.71 loss=(-0.004,4.877) kl=0.0007 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_016000.safetensors +update=16000 collisions=17.68 chaser_return=18.88 explorer_return=18.34 vision=(0.932,0.932) new_fields=(49.8,55.5) distance=1.68 loss=(-0.004,4.763) kl=0.0012 +update=16500 collisions=21.10 chaser_return=22.72 explorer_return=19.21 vision=(0.935,0.935) new_fields=(51.7,58.7) distance=1.96 loss=(-0.004,4.645) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_017000.safetensors +update=17000 collisions=18.73 chaser_return=20.13 explorer_return=22.95 vision=(0.930,0.930) new_fields=(50.9,59.8) distance=1.79 loss=(-0.003,4.765) kl=0.0005 +update=17500 collisions=18.80 chaser_return=20.43 explorer_return=26.21 vision=(0.888,0.888) new_fields=(52.6,61.2) distance=1.85 loss=(-0.003,4.952) kl=0.0014 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_018000.safetensors +update=18000 collisions=18.33 chaser_return=19.10 explorer_return=23.00 vision=(0.924,0.924) new_fields=(48.0,58.7) distance=1.95 loss=(-0.005,5.410) kl=0.0014 +update=18500 collisions=19.33 chaser_return=20.35 explorer_return=16.84 vision=(0.943,0.943) new_fields=(49.3,55.8) distance=1.61 loss=(-0.003,4.171) kl=0.0009 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors +update=19000 collisions=18.02 chaser_return=18.94 explorer_return=22.70 vision=(0.940,0.940) new_fields=(49.8,58.8) distance=1.75 loss=(-0.005,5.021) kl=0.0007 +update=19500 collisions=18.23 chaser_return=19.14 explorer_return=21.96 vision=(0.937,0.937) new_fields=(50.2,57.9) distance=1.57 loss=(-0.002,4.085) kl=0.0001 +update=20000 collisions=21.25 chaser_return=22.77 explorer_return=21.69 vision=(0.955,0.955) new_fields=(51.6,58.4) distance=1.54 loss=(-0.003,4.631) kl=-0.0003 +saved_checkpoint=runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors +MRR_JOB_RESULT={"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "elapsed_seconds": 7585.706820899039, "metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "task": "social", "unit_id": "social/seed_0009"} +=== exited 2026-07-02T11:50:30.964111+00:00 returncode=0 === diff --git a/manifest.json b/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..a9c1ccc4bcbef0d78db038cf346eb3f59d00c84d --- /dev/null +++ b/manifest.json @@ -0,0 +1,423 @@ +{ + "config": { + "architecture": "mlp", + "batch_size": 40, + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "concurrency": 10, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "finite_guard": true, + "fused_agent_rollout": false, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "max_steps": 100, + "ppo_epochs": 4, + "preset": "paper_text", + "recurrent_l2_coef": 0.3, + "status_every_seconds": 1800.0, + "subspace_metric_period": 10, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 + }, + "config_sha256": "1e164a3dab3a7ecde08c0595611a60ed85489e06861f477e46efeb1be25d94c3", + "created_at": "2026-07-02T07:36:16.447422+00:00", + "experiment": "mouse-run-run-2-mlp", + "experiment_spec": "experiments/paper_marl_2026/SPEC.md", + "experiment_spec_sha256": "e878bd03a06735c93c8caab4fc769ce63f026a538dfccbc87b588d432e48761e", + "initial_attempts": [ + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0000_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", + "seed": 0, + "task": "social", + "unit_id": "social/seed_0000" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0000_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", + "seed": 0, + "task": "non_social", + "unit_id": "non_social/seed_0000" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0001_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", + "seed": 1, + "task": "social", + "unit_id": "social/seed_0001" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0001_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", + "seed": 1, + "task": "non_social", + "unit_id": "non_social/seed_0001" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0002_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", + "seed": 2, + "task": "social", + "unit_id": "social/seed_0002" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0002_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", + "seed": 2, + "task": "non_social", + "unit_id": "non_social/seed_0002" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0003_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", + "seed": 3, + "task": "social", + "unit_id": "social/seed_0003" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0003_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", + "seed": 3, + "task": "non_social", + "unit_id": "non_social/seed_0003" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0004_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", + "seed": 4, + "task": "social", + "unit_id": "social/seed_0004" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0004_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", + "seed": 4, + "task": "non_social", + "unit_id": "non_social/seed_0004" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0005_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", + "seed": 5, + "task": "social", + "unit_id": "social/seed_0005" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0005_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", + "seed": 5, + "task": "non_social", + "unit_id": "non_social/seed_0005" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0006_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", + "seed": 6, + "task": "social", + "unit_id": "social/seed_0006" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0006_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", + "seed": 6, + "task": "non_social", + "unit_id": "non_social/seed_0006" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0007_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", + "seed": 7, + "task": "social", + "unit_id": "social/seed_0007" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0007_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", + "seed": 7, + "task": "non_social", + "unit_id": "non_social/seed_0007" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0008_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", + "seed": 8, + "task": "social", + "unit_id": "social/seed_0008" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0008_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", + "seed": 8, + "task": "non_social", + "unit_id": "non_social/seed_0008" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0009_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", + "seed": 9, + "task": "social", + "unit_id": "social/seed_0009" + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0009_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", + "seed": 9, + "task": "non_social", + "unit_id": "non_social/seed_0009" + } + ], + "max_attempts": 2, + "provenance": { + "command": "scripts/run_local_experiment.py --config experiments/paper_marl_2026/configs/arch_mlp.json", + "created_at": "2026-07-02T07:36:16.447522+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "schema_version": 2, + "total_environment_steps": 1600000000, + "total_units": 20, + "units": [ + { + "seed": 0, + "task": "social", + "unit_id": "social/seed_0000" + }, + { + "seed": 0, + "task": "non_social", + "unit_id": "non_social/seed_0000" + }, + { + "seed": 1, + "task": "social", + "unit_id": "social/seed_0001" + }, + { + "seed": 1, + "task": "non_social", + "unit_id": "non_social/seed_0001" + }, + { + "seed": 2, + "task": "social", + "unit_id": "social/seed_0002" + }, + { + "seed": 2, + "task": "non_social", + "unit_id": "non_social/seed_0002" + }, + { + "seed": 3, + "task": "social", + "unit_id": "social/seed_0003" + }, + { + "seed": 3, + "task": "non_social", + "unit_id": "non_social/seed_0003" + }, + { + "seed": 4, + "task": "social", + "unit_id": "social/seed_0004" + }, + { + "seed": 4, + "task": "non_social", + "unit_id": "non_social/seed_0004" + }, + { + "seed": 5, + "task": "social", + "unit_id": "social/seed_0005" + }, + { + "seed": 5, + "task": "non_social", + "unit_id": "non_social/seed_0005" + }, + { + "seed": 6, + "task": "social", + "unit_id": "social/seed_0006" + }, + { + "seed": 6, + "task": "non_social", + "unit_id": "non_social/seed_0006" + }, + { + "seed": 7, + "task": "social", + "unit_id": "social/seed_0007" + }, + { + "seed": 7, + "task": "non_social", + "unit_id": "non_social/seed_0007" + }, + { + "seed": 8, + "task": "social", + "unit_id": "social/seed_0008" + }, + { + "seed": 8, + "task": "non_social", + "unit_id": "non_social/seed_0008" + }, + { + "seed": 9, + "task": "social", + "unit_id": "social/seed_0009" + }, + { + "seed": 9, + "task": "non_social", + "unit_id": "non_social/seed_0009" + } + ] +} diff --git a/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..42ccaee0cc6649de934e41e961c872d5566d6951 --- /dev/null +++ b/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f8e5409f712499d04d8944b4ecd7070e6a1688c886ed4463cf04bf85fc9707a3 +size 11460376 diff --git a/non_social/seed_0000/attempt_01/config.json b/non_social/seed_0000/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..214865e4f6aced3fdd32e99df4263a643afeed99 --- /dev/null +++ b/non_social/seed_0000/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", + "seed": 0, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0000/attempt_01/metrics.jsonl b/non_social/seed_0000/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6143fca14cb2ebf6012cd858424ab9d11cb510de --- /dev/null +++ b/non_social/seed_0000/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005362351288850833, "elapsed_seconds": 1.9304464639863, "env_steps_per_second": 2072.059533699862, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.72416634257278, "eta_seconds": 38606.998833262005, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -5.173608782563233e-06, "chaser_grad_norm": 0.34016749262809753, "chaser_new_fields": 35.775001525878906, "chaser_partner_vision": 0.0, "chaser_return": -2.9100003242492676, "chaser_subspace_norm": 0.0029390037525445223, "collisions_per_episode": 1.7000000476837158, "entropy": 2.7725364565849304, "explorer_grad_norm": 0.1394939087331295, "explorer_new_fields": 33.70000076293945, "explorer_partner_vision": 0.0, "explorer_return": 0.10000000149011612, "explorer_subspace_norm": 0.0024945889599621296, "final_distance": 5.466697692871094, "max_param_abs": 0.15428763628005981, "policy_loss": -0.000497843058607117, "value_loss": 3.074083089828491}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.653523+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.6535435, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05329365353998987, "elapsed_seconds": 191.85715274396352, "env_steps_per_second": 10424.422396537033, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.078452488059605, "eta_seconds": 7482.428957014578, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 6.691479853282445e-05, "chaser_grad_norm": 0.23669546842575073, "chaser_new_fields": 74.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.700000762939453, "chaser_subspace_norm": 16.239971160888672, "collisions_per_episode": 3.375, "entropy": 0.4663320481777191, "explorer_grad_norm": 0.3577277585864067, "explorer_new_fields": 67.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 49.9375, "explorer_subspace_norm": 8.448774337768555, "final_distance": 4.963988780975342, "max_param_abs": 0.3258746862411499, "policy_loss": -0.0018085034082275797, "value_loss": 4.557863235473633}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:29.580229+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977969.5802422, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10710371320084151, "elapsed_seconds": 385.57336752302945, "env_steps_per_second": 10374.160502050467, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0349705508159883, "eta_seconds": 7325.893982937559, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0007281143502567144, "chaser_grad_norm": 0.17577309906482697, "chaser_new_fields": 79.5, "chaser_partner_vision": 0.0, "chaser_return": 5.810000896453857, "chaser_subspace_norm": 26.5023136138916, "collisions_per_episode": 1.899999976158142, "entropy": 0.3493778705596924, "explorer_grad_norm": 0.6742401272058487, "explorer_new_fields": 67.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 50.42499923706055, "explorer_subspace_norm": 8.997306823730469, "final_distance": 5.553553581237793, "max_param_abs": 0.3305240869522095, "policy_loss": -0.0016281890529228349, "value_loss": 4.399340271949768}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:43.296439+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978163.2964523, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.1595268888483406, "elapsed_seconds": 574.2967998540262, "env_steps_per_second": 10447.559522402127, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9674982957962008, "eta_seconds": 7082.993864866323, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0006018214631288155, "chaser_grad_norm": 0.21672200597822666, "chaser_new_fields": 80.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.075000762939453, "chaser_subspace_norm": 36.892730712890625, "collisions_per_episode": 3.049999952316284, "entropy": 0.3051275759935379, "explorer_grad_norm": 0.4998535215854645, "explorer_new_fields": 69.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 53.650001525878906, "explorer_subspace_norm": 8.920053482055664, "final_distance": 5.113457679748535, "max_param_abs": 0.349557101726532, "policy_loss": -0.0010127461082305445, "value_loss": 3.9467450380325317}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:52.019871+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978352.0198798, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21280761047722382, "elapsed_seconds": 766.1073977180058, "env_steps_per_second": 10442.400143673714, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.915268494295015, "eta_seconds": 6894.966579462053, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0008088869087739536, "chaser_grad_norm": 0.07538860198110342, "chaser_new_fields": 79.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.820001125335693, "chaser_subspace_norm": 42.316566467285156, "collisions_per_episode": 2.125, "entropy": 0.30866675823926926, "explorer_grad_norm": 1.8579207435250282, "explorer_new_fields": 67.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 50.42499923706055, "explorer_subspace_norm": 10.101943969726562, "final_distance": 4.400704383850098, "max_param_abs": 0.4545920789241791, "policy_loss": -0.0013379232555177367, "value_loss": 4.046706676483154}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:03.830470+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978543.8304827, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2660434453658268, "elapsed_seconds": 957.7564033169765, "env_steps_per_second": 10441.068277243798, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8623041175607877, "eta_seconds": 6704.2948232188355, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0006254628758133318, "chaser_grad_norm": 0.2807302428409457, "chaser_new_fields": 76.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.1800007820129395, "chaser_subspace_norm": 51.4793701171875, "collisions_per_episode": 2.299999952316284, "entropy": 0.23320892453193665, "explorer_grad_norm": 1.430855542421341, "explorer_new_fields": 70.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 55.11249923706055, "explorer_subspace_norm": 10.984577178955078, "final_distance": 5.455503940582275, "max_param_abs": 0.511533260345459, "policy_loss": -0.0011108804219404078, "value_loss": 3.958456575870514}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:15.479474+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978735.4794846, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3191621872458361, "elapsed_seconds": 1148.98387408501, "env_steps_per_second": 10444.010808729727, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8085857277264048, "eta_seconds": 6510.908619815057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00035122274371035545, "chaser_grad_norm": 0.10724705085158348, "chaser_new_fields": 77.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.355000972747803, "chaser_subspace_norm": 54.21003341674805, "collisions_per_episode": 2.375, "entropy": 0.26491987705230713, "explorer_grad_norm": 0.5290967896580696, "explorer_new_fields": 67.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 50.57500076293945, "explorer_subspace_norm": 10.944963455200195, "final_distance": 5.601850509643555, "max_param_abs": 0.5453172922134399, "policy_loss": -0.0015410876681567487, "value_loss": 3.992495119571686}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:26.706946+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978926.7069602, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.37103237310637877, "elapsed_seconds": 1335.7165431829635, "env_steps_per_second": 10481.265708245639, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7491526160729283, "eta_seconds": 6296.949417862542, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0001249361482511535, "chaser_grad_norm": 0.12887633964419365, "chaser_new_fields": 80.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.02000093460083, "chaser_subspace_norm": 57.67909240722656, "collisions_per_episode": 1.899999976158142, "entropy": 0.2018074169754982, "explorer_grad_norm": 0.512401819229126, "explorer_new_fields": 70.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 54.36249923706055, "explorer_subspace_norm": 11.544838905334473, "final_distance": 4.9441657066345215, "max_param_abs": 0.5763495564460754, "policy_loss": -0.00044214877009096654, "value_loss": 3.771480619907379}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:33.439614+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979113.4396262, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4231054240669538, "elapsed_seconds": 1523.1795266410336, "env_steps_per_second": 10504.342869736265, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6924216962678151, "eta_seconds": 6092.718106564134, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00014633760644949811, "chaser_grad_norm": 0.054581169970333576, "chaser_new_fields": 79.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.875001430511475, "chaser_subspace_norm": 62.2651252746582, "collisions_per_episode": 1.649999976158142, "entropy": 0.19518577307462692, "explorer_grad_norm": 0.6297692358493805, "explorer_new_fields": 71.875, "explorer_partner_vision": 0.0, "explorer_return": 57.17499923706055, "explorer_subspace_norm": 12.46562671661377, "final_distance": 5.8548688888549805, "max_param_abs": 0.5877197980880737, "policy_loss": -0.0008621873332899099, "value_loss": 3.66826194524765}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:40.902600+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979300.9026148, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4754179403783443, "elapsed_seconds": 1711.5045853620395, "env_steps_per_second": 10517.062094924162, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.637550683525408, "eta_seconds": 5895.182460691469, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.000319335286664002, "chaser_grad_norm": 0.12068943493068218, "chaser_new_fields": 81.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.285000801086426, "chaser_subspace_norm": 70.37218475341797, "collisions_per_episode": 1.625, "entropy": 0.179047342389822, "explorer_grad_norm": 0.524848110973835, "explorer_new_fields": 69.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 54.13750076293945, "explorer_subspace_norm": 13.666532516479492, "final_distance": 5.531441688537598, "max_param_abs": 0.5752404928207397, "policy_loss": -0.0008478180620574705, "value_loss": 3.7342492938041687}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:04:49.227656+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979489.227666, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5281459557525038, "elapsed_seconds": 1901.3254407090135, "env_steps_per_second": 10518.977746672292, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5844378672575112, "eta_seconds": 5703.9763221270405, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0008280922713789496, "chaser_grad_norm": 0.11795545369386673, "chaser_new_fields": 77.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.380000591278076, "chaser_subspace_norm": 69.02265167236328, "collisions_per_episode": 2.1500000953674316, "entropy": 0.14308607205748558, "explorer_grad_norm": 0.48360154777765274, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.087501525878906, "explorer_subspace_norm": 13.96146011352539, "final_distance": 5.5770416259765625, "max_param_abs": 0.647881269454956, "policy_loss": -0.00018160262168720465, "value_loss": 3.802269697189331}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:07:59.048512+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979679.048526, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5800792278741654, "elapsed_seconds": 2088.2852203469956, "env_steps_per_second": 10534.959394265317, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5292997825773451, "eta_seconds": 5505.479217278443, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": -4.474588249175904e-05, "chaser_grad_norm": 0.13463049568235874, "chaser_new_fields": 78.75, "chaser_partner_vision": 0.0, "chaser_return": 5.690001010894775, "chaser_subspace_norm": 68.8963623046875, "collisions_per_episode": 1.625, "entropy": 0.13056070357561111, "explorer_grad_norm": 0.8507837206125259, "explorer_new_fields": 72.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 58.525001525878906, "explorer_subspace_norm": 15.163471221923828, "final_distance": 5.202518463134766, "max_param_abs": 0.7021549940109253, "policy_loss": -0.0006455526853557103, "value_loss": 3.8411065340042114}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:06.008292+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979866.008303, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6321262312230667, "elapsed_seconds": 2275.65443240304, "env_steps_per_second": 10546.416739845925, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4749612061871558, "eta_seconds": 5309.860342273761, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00040526999702991673, "chaser_grad_norm": 0.08878211490809917, "chaser_new_fields": 78.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.560000896453857, "chaser_subspace_norm": 68.70912170410156, "collisions_per_episode": 1.850000023841858, "entropy": 0.13891512528061867, "explorer_grad_norm": 0.6218316107988358, "explorer_new_fields": 72.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.70000076293945, "explorer_subspace_norm": 16.02243423461914, "final_distance": 4.691096305847168, "max_param_abs": 0.7957910895347595, "policy_loss": -0.0008608128022196126, "value_loss": 3.6975609064102173}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:13.377505+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980053.3775182, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6850991590336263, "elapsed_seconds": 2466.3569725210546, "env_steps_per_second": 10541.864089294173, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4228982533775314, "eta_seconds": 5122.433712159113, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0002404470562145239, "chaser_grad_norm": 0.09046696685254574, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.440001010894775, "chaser_subspace_norm": 77.56379699707031, "collisions_per_episode": 1.725000023841858, "entropy": 0.11232517659664154, "explorer_grad_norm": 0.6239490285515785, "explorer_new_fields": 72.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 58.60000228881836, "explorer_subspace_norm": 17.372209548950195, "final_distance": 5.052657604217529, "max_param_abs": 0.849467933177948, "policy_loss": -0.0007943277134287996, "value_loss": 3.6960580348968506}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:24.080043+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980244.080054, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7369473368947123, "elapsed_seconds": 2653.010412820964, "env_steps_per_second": 10554.048285934698, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3686164828044656, "eta_seconds": 4927.019338096076, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -1.6540823666894022e-05, "chaser_grad_norm": 0.11463696509599686, "chaser_new_fields": 82.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.4250006675720215, "chaser_subspace_norm": 78.05644226074219, "collisions_per_episode": 1.600000023841858, "entropy": 0.11654260009527206, "explorer_grad_norm": 0.8265438824892044, "explorer_new_fields": 77.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 65.5, "explorer_subspace_norm": 16.70997428894043, "final_distance": 4.866159915924072, "max_param_abs": 0.862048327922821, "policy_loss": -0.0003952661583070949, "value_loss": 3.6717289686203003}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:20:30.733484+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980430.733498, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7888488838738865, "elapsed_seconds": 2839.8559819459915, "env_steps_per_second": 10563.915984022087, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3147481397898109, "eta_seconds": 4733.0933032433195, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00047657695344544004, "chaser_grad_norm": 0.06467235274612904, "chaser_new_fields": 80.25, "chaser_partner_vision": 0.0, "chaser_return": 5.950000762939453, "chaser_subspace_norm": 79.99512481689453, "collisions_per_episode": 1.774999976158142, "entropy": 0.09236839413642883, "explorer_grad_norm": 1.273687481880188, "explorer_new_fields": 77.25, "explorer_partner_vision": 0.0, "explorer_return": 65.05000305175781, "explorer_subspace_norm": 18.648405075073242, "final_distance": 5.496772766113281, "max_param_abs": 0.8750737905502319, "policy_loss": -0.0006252801441561928, "value_loss": 3.7000213265419006}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:23:37.579052+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980617.579064, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.840673852563066, "elapsed_seconds": 3026.4258692270378, "env_steps_per_second": 10573.528440058219, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.261010778844599, "eta_seconds": 4539.638803840557, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 1.1703377388594394e-06, "chaser_grad_norm": 0.20380252599716187, "chaser_new_fields": 80.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.945001125335693, "chaser_subspace_norm": 82.45390319824219, "collisions_per_episode": 1.649999976158142, "entropy": 0.10570455901324749, "explorer_grad_norm": 1.6526433825492859, "explorer_new_fields": 75.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 63.25, "explorer_subspace_norm": 18.309865951538086, "final_distance": 5.806778430938721, "max_param_abs": 0.9599581360816956, "policy_loss": -0.0004156225741596131, "value_loss": 3.573492467403412}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:26:44.148942+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980804.1489553, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8938823972936047, "elapsed_seconds": 3217.976630256977, "env_steps_per_second": 10565.645405972036, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2093703022207591, "eta_seconds": 4353.733087994733, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0004711518696320638, "chaser_grad_norm": 0.043468572199344635, "chaser_new_fields": 81.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.185000896453857, "chaser_subspace_norm": 75.3616714477539, "collisions_per_episode": 1.9250000715255737, "entropy": 0.11739776842296124, "explorer_grad_norm": 0.8637182414531708, "explorer_new_fields": 76.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 64.5625, "explorer_subspace_norm": 19.596525192260742, "final_distance": 5.308167934417725, "max_param_abs": 1.043190836906433, "policy_loss": -0.0012657396960946699, "value_loss": 3.582657754421234}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:29:55.699701+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980995.6997137, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9460653386549933, "elapsed_seconds": 3405.835219157976, "env_steps_per_second": 10570.094465374714, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1563020805783253, "eta_seconds": 4162.687490081971, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0004890187930004786, "chaser_grad_norm": 0.12140505574643612, "chaser_new_fields": 79.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.7400007247924805, "chaser_subspace_norm": 82.12152099609375, "collisions_per_episode": 1.8250000476837158, "entropy": 0.10391798242926598, "explorer_grad_norm": 1.8681467473506927, "explorer_new_fields": 78.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 66.32500457763672, "explorer_subspace_norm": 18.82545280456543, "final_distance": 4.985391139984131, "max_param_abs": 1.093315839767456, "policy_loss": -0.0009176402622799174, "value_loss": 3.6649279594421387}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:03.558290+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981183.5583043, "update": 9000} +{"device": "cuda", "elapsed_hours": 0.9989255789939003, "elapsed_seconds": 3596.1320843780413, "env_steps_per_second": 10566.908864409017, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1040756399406266, "eta_seconds": 3974.672303786256, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00019615038952985886, "chaser_grad_norm": 0.14012088812887669, "chaser_new_fields": 81.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.160000801086426, "chaser_subspace_norm": 80.92264556884766, "collisions_per_episode": 1.8250000476837158, "entropy": 0.09415227919816971, "explorer_grad_norm": 0.38225405663251877, "explorer_new_fields": 77.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 64.9375, "explorer_subspace_norm": 20.194814682006836, "final_distance": 4.540725231170654, "max_param_abs": 1.180087924003601, "policy_loss": -0.000374201442522093, "value_loss": 3.4838927388191223}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:13.855155+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981373.8551664, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0525457070244333, "elapsed_seconds": 3789.16454528796, "env_steps_per_second": 10556.4167303693, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0525457070244333, "eta_seconds": 3789.16454528796, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -0.0001693130862814201, "chaser_grad_norm": 0.07559763826429844, "chaser_new_fields": 81.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.155000686645508, "chaser_subspace_norm": 85.33003997802734, "collisions_per_episode": 2.25, "entropy": 0.08448796719312668, "explorer_grad_norm": 1.4887766242027283, "explorer_new_fields": 77.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 65.2750015258789, "explorer_subspace_norm": 21.593774795532227, "final_distance": 5.322676181793213, "max_param_abs": 1.2174458503723145, "policy_loss": -1.3907546148272942e-05, "value_loss": 3.5226200222969055}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:39:26.887618+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981566.8876326, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1047508845430436, "elapsed_seconds": 3977.103184354957, "env_steps_per_second": 10560.450170168753, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9995365145865633, "eta_seconds": 3598.331452511628, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -1.5043517249679894e-06, "chaser_grad_norm": 0.05930179636925459, "chaser_new_fields": 82.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.415000915527344, "chaser_subspace_norm": 86.52581787109375, "collisions_per_episode": 2.0, "entropy": 0.06573069468140602, "explorer_grad_norm": 0.4400622770190239, "explorer_new_fields": 77.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 65.20000457763672, "explorer_subspace_norm": 22.82740592956543, "final_distance": 5.1441144943237305, "max_param_abs": 1.2872856855392456, "policy_loss": -0.0009458466398069021, "value_loss": 3.4636882543563843}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:42:34.826255+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981754.8262658, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.158430229554716, "elapsed_seconds": 4170.3488263969775, "env_steps_per_second": 10550.676174015478, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9478065514538585, "eta_seconds": 3412.1035852338905, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 7.785861745286671e-05, "chaser_grad_norm": 0.04884711094200611, "chaser_new_fields": 83.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.520001411437988, "chaser_subspace_norm": 87.99630737304688, "collisions_per_episode": 1.524999976158142, "entropy": 0.06133109796792269, "explorer_grad_norm": 0.802131175994873, "explorer_new_fields": 78.25, "explorer_partner_vision": 0.0, "explorer_return": 66.5875015258789, "explorer_subspace_norm": 23.026655197143555, "final_distance": 4.931851863861084, "max_param_abs": 1.2872856855392456, "policy_loss": -0.00041549857851563843, "value_loss": 3.494303584098816}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:45:48.071898+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981948.0719104, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2102489595363861, "elapsed_seconds": 4356.89625433099, "env_steps_per_second": 10557.974602740085, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8945318396573289, "eta_seconds": 3220.3146227663838, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 8.80808145280465e-05, "chaser_grad_norm": 0.07423081062734127, "chaser_new_fields": 80.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.070000648498535, "chaser_subspace_norm": 80.28471374511719, "collisions_per_episode": 1.399999976158142, "entropy": 0.061599936336278915, "explorer_grad_norm": 1.113692745566368, "explorer_new_fields": 79.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.32500457763672, "explorer_subspace_norm": 24.018014907836914, "final_distance": 5.806375503540039, "max_param_abs": 1.3213297128677368, "policy_loss": -0.0020681732161946798, "value_loss": 3.4564942121505737}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:48:54.619325+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982134.6193366, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2624173470558404, "elapsed_seconds": 4544.702449401026, "env_steps_per_second": 10561.747558704577, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8416115647038938, "eta_seconds": 3029.8016329340176, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0008321205401692622, "chaser_grad_norm": 0.12845297157764435, "chaser_new_fields": 75.875, "chaser_partner_vision": 0.0, "chaser_return": 5.105000972747803, "chaser_subspace_norm": 81.34334564208984, "collisions_per_episode": 1.649999976158142, "entropy": 0.09196563623845577, "explorer_grad_norm": 0.29978178441524506, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.17500305175781, "explorer_subspace_norm": 24.92875862121582, "final_distance": 5.172975063323975, "max_param_abs": 1.3213297128677368, "policy_loss": -0.0010934120214081133, "value_loss": 3.763103723526001}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:02.425521+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982322.425536, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3171340958397357, "elapsed_seconds": 4741.6827450230485, "env_steps_per_second": 10544.779709794135, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7902804575038415, "eta_seconds": 2845.009647013829, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00044572072633515536, "chaser_grad_norm": 0.07732724957168102, "chaser_new_fields": 84.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.785000801086426, "chaser_subspace_norm": 78.33721160888672, "collisions_per_episode": 1.350000023841858, "entropy": 0.08976677805185318, "explorer_grad_norm": 1.0155585557222366, "explorer_new_fields": 79.125, "explorer_partner_vision": 0.0, "explorer_return": 68.125, "explorer_subspace_norm": 23.136762619018555, "final_distance": 5.725672245025635, "max_param_abs": 1.3213297128677368, "policy_loss": -0.001024966465664079, "value_loss": 3.352021098136902}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:19.405818+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982519.4058306, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3712515359733435, "elapsed_seconds": 4936.5055295040365, "env_steps_per_second": 10533.767194062955, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7383662116779541, "eta_seconds": 2658.118362040635, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0004590160541175642, "chaser_grad_norm": 0.06538167875260115, "chaser_new_fields": 83.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.54500150680542, "chaser_subspace_norm": 78.45780181884766, "collisions_per_episode": 1.399999976158142, "entropy": 0.09547889046370983, "explorer_grad_norm": 0.5435245260596275, "explorer_new_fields": 78.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 67.63750457763672, "explorer_subspace_norm": 24.118432998657227, "final_distance": 5.541812419891357, "max_param_abs": 1.3252073526382446, "policy_loss": -0.0007060517087804641, "value_loss": 3.4721049070358276}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:58:34.228603+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982714.2286181, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4240544896291816, "elapsed_seconds": 5126.596162665053, "env_steps_per_second": 10533.304806268996, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6856558653770133, "eta_seconds": 2468.361115357248, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0008695930212665459, "chaser_grad_norm": 0.0758396191522479, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.42000150680542, "chaser_subspace_norm": 74.20014190673828, "collisions_per_episode": 1.725000023841858, "entropy": 0.050692721270024776, "explorer_grad_norm": 1.3435713052749634, "explorer_new_fields": 80.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.70000457763672, "explorer_subspace_norm": 25.260629653930664, "final_distance": 5.570320129394531, "max_param_abs": 1.3920022249221802, "policy_loss": -0.0019914439972481546, "value_loss": 3.334362030029297}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:01:44.319234+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982904.3192446, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4773909531022138, "elapsed_seconds": 5318.6074311679695, "env_steps_per_second": 10529.071890478363, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6331675513295202, "eta_seconds": 2279.4031847862725, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0002321924548440088, "chaser_grad_norm": 0.06448401790112257, "chaser_new_fields": 82.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.415000915527344, "chaser_subspace_norm": 81.34850311279297, "collisions_per_episode": 1.8250000476837158, "entropy": 0.06315584108233452, "explorer_grad_norm": 0.8479587733745575, "explorer_new_fields": 78.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.625, "explorer_subspace_norm": 27.18760871887207, "final_distance": 5.711249351501465, "max_param_abs": 1.4050794839859009, "policy_loss": -0.0011196303819325237, "value_loss": 3.3451501727104187}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:04:56.330503+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983096.330517, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5296515264125061, "elapsed_seconds": 5506.745495085022, "env_steps_per_second": 10532.536877138626, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5802126479495713, "eta_seconds": 2088.7655326184567, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -6.950403765165585e-05, "chaser_grad_norm": 0.062295411713421345, "chaser_new_fields": 82.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 79.73140716552734, "collisions_per_episode": 1.7000000476837158, "entropy": 0.05496581457555294, "explorer_grad_norm": 1.555381953716278, "explorer_new_fields": 79.125, "explorer_partner_vision": 0.0, "explorer_return": 68.01250457763672, "explorer_subspace_norm": 26.886980056762695, "final_distance": 5.146921634674072, "max_param_abs": 1.4050794839859009, "policy_loss": -0.00048470002552480906, "value_loss": 3.3966342210769653}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:04.468566+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983284.4685764, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5815020676722311, "elapsed_seconds": 5693.407443620032, "env_steps_per_second": 10538.50450616095, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5271673558907437, "eta_seconds": 1897.8024812066772, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00032535001055494783, "chaser_grad_norm": 0.0946476785466075, "chaser_new_fields": 84.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.770000457763672, "chaser_subspace_norm": 81.79399871826172, "collisions_per_episode": 1.399999976158142, "entropy": 0.08795448578894138, "explorer_grad_norm": 1.0476092919707298, "explorer_new_fields": 78.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 67.5250015258789, "explorer_subspace_norm": 25.196977615356445, "final_distance": 5.499534606933594, "max_param_abs": 1.420541524887085, "policy_loss": -0.0003041045347949023, "value_loss": 3.2325546741485596}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:11.130515+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983471.1305299, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6352512755986148, "elapsed_seconds": 5886.904592155013, "env_steps_per_second": 10531.850657580255, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4747503703350817, "eta_seconds": 1709.1013332062942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": -8.603196387602097e-05, "chaser_grad_norm": 0.05303043685853481, "chaser_new_fields": 83.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.555000305175781, "chaser_subspace_norm": 80.69808197021484, "collisions_per_episode": 1.1750000715255737, "entropy": 0.08621848374605179, "explorer_grad_norm": 0.94834154099226, "explorer_new_fields": 76.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 63.85000228881836, "explorer_subspace_norm": 24.56145477294922, "final_distance": 5.3362135887146, "max_param_abs": 1.435640811920166, "policy_loss": -0.0006088202145186905, "value_loss": 3.4417847990989685}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:14:24.627665+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983664.6276793, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6885142894677767, "elapsed_seconds": 6078.651442083996, "env_steps_per_second": 10528.6510683788, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4221285723669442, "eta_seconds": 1519.662860520999, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 8.177443310553034e-05, "chaser_grad_norm": 0.026209967210888863, "chaser_new_fields": 82.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.415000915527344, "chaser_subspace_norm": 80.78495025634766, "collisions_per_episode": 1.7000000476837158, "entropy": 0.0644609984010458, "explorer_grad_norm": 1.3456061631441116, "explorer_new_fields": 80.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.9625015258789, "explorer_subspace_norm": 26.441322326660156, "final_distance": 5.649790287017822, "max_param_abs": 1.486730933189392, "policy_loss": -0.0005736820357267192, "value_loss": 3.2932186126708984}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:17:36.374514+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983856.3745298, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7405432201391604, "elapsed_seconds": 6265.955592500977, "env_steps_per_second": 10533.110078052267, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3692061376052764, "eta_seconds": 1329.142095378995, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 1.3870245280732385e-05, "chaser_grad_norm": 0.03339986642822623, "chaser_new_fields": 83.625, "chaser_partner_vision": 0.0, "chaser_return": 6.625000953674316, "chaser_subspace_norm": 81.69694519042969, "collisions_per_episode": 1.5750000476837158, "entropy": 0.05169977340847254, "explorer_grad_norm": 1.3918738961219788, "explorer_new_fields": 81.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 71.5, "explorer_subspace_norm": 25.642988204956055, "final_distance": 5.413881301879883, "max_param_abs": 1.4971561431884766, "policy_loss": -0.0004798526813694082, "value_loss": 3.2801648378372192}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:20:43.678663+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984043.6786747, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7924729026744415, "elapsed_seconds": 6452.902449627989, "env_steps_per_second": 10537.893689051538, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3163187475307838, "eta_seconds": 1138.7474911108218, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00021581246380400487, "chaser_grad_norm": 0.023911027936264873, "chaser_new_fields": 82.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.355000972747803, "chaser_subspace_norm": 77.47178649902344, "collisions_per_episode": 1.475000023841858, "entropy": 0.05338282324373722, "explorer_grad_norm": 0.8776888698339462, "explorer_new_fields": 80.5, "explorer_partner_vision": 0.0, "explorer_return": 70.0374984741211, "explorer_subspace_norm": 25.362728118896484, "final_distance": 5.1118950843811035, "max_param_abs": 1.5214029550552368, "policy_loss": -0.00040889874164840023, "value_loss": 3.3093971610069275}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:23:50.625521+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984230.6255345, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8450617680116557, "elapsed_seconds": 6642.22236484196, "env_steps_per_second": 10538.641459900226, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26358025257309364, "eta_seconds": 948.8889092631371, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0002657267602084179, "chaser_grad_norm": 0.05063078086823225, "chaser_new_fields": 82.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.455001354217529, "chaser_subspace_norm": 75.80194854736328, "collisions_per_episode": 1.350000023841858, "entropy": 0.05619639251381159, "explorer_grad_norm": 0.7586047500371933, "explorer_new_fields": 81.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 71.80000305175781, "explorer_subspace_norm": 25.632326126098633, "final_distance": 5.240739345550537, "max_param_abs": 1.5486178398132324, "policy_loss": -0.0004507996744940179, "value_loss": 3.1661828756332397}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:26:59.945437+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984419.9454508, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.8987893208727797, "elapsed_seconds": 6835.641555142007, "env_steps_per_second": 10533.02742971347, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21097659120808665, "eta_seconds": 759.5157283491119, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.000476315018807405, "chaser_grad_norm": 0.0847174865193665, "chaser_new_fields": 82.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.40500020980835, "chaser_subspace_norm": 73.47077178955078, "collisions_per_episode": 1.1750000715255737, "entropy": 0.05018320120871067, "explorer_grad_norm": 1.9407499842345715, "explorer_new_fields": 79.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 68.8375015258789, "explorer_subspace_norm": 24.79322052001953, "final_distance": 5.659545421600342, "max_param_abs": 1.5856611728668213, "policy_loss": -0.0004987320494791447, "value_loss": 3.22615909576416}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:30:13.364629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984613.3646438, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9517543001891722, "elapsed_seconds": 7026.31548068102, "env_steps_per_second": 10531.83566884583, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15825034866398693, "eta_seconds": 569.701255190353, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00022815836740864605, "chaser_grad_norm": 0.0501972446218133, "chaser_new_fields": 81.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.215001106262207, "chaser_subspace_norm": 81.19401550292969, "collisions_per_episode": 1.524999976158142, "entropy": 0.059275402687489986, "explorer_grad_norm": 0.6187672689557076, "explorer_new_fields": 79.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.55000305175781, "explorer_subspace_norm": 26.134788513183594, "final_distance": 6.221502780914307, "max_param_abs": 1.6402511596679688, "policy_loss": -0.0005101506211815376, "value_loss": 3.2488821148872375}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:33:24.038552+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984804.0385628, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.00550099451415, "elapsed_seconds": 7219.80358025094, "env_steps_per_second": 10526.602165173925, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10555268392179737, "eta_seconds": 379.9896621184705, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -1.0564466708995468e-05, "chaser_grad_norm": 0.13344785012304783, "chaser_new_fields": 81.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.220000743865967, "chaser_subspace_norm": 77.87812042236328, "collisions_per_episode": 1.3000000715255737, "entropy": 0.057777862064540386, "explorer_grad_norm": 0.9379898607730865, "explorer_new_fields": 77.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 65.98750305175781, "explorer_subspace_norm": 25.909114837646484, "final_distance": 5.9693779945373535, "max_param_abs": 1.6938241720199585, "policy_loss": -0.0004634228043762789, "value_loss": 3.289840877056122}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:36:37.526654+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984997.5266697, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.058184965768046, "elapsed_seconds": 7409.465876764967, "env_steps_per_second": 10527.074595835164, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05277397348123196, "eta_seconds": 189.98630453243504, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 8.450402172603677e-05, "chaser_grad_norm": 0.0522395009174943, "chaser_new_fields": 85.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.895001411437988, "chaser_subspace_norm": 85.82374572753906, "collisions_per_episode": 1.524999976158142, "entropy": 0.0807732418179512, "explorer_grad_norm": 0.4299086779356003, "explorer_new_fields": 76.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 64.11250305175781, "explorer_subspace_norm": 25.97504425048828, "final_distance": 6.12801456451416, "max_param_abs": 1.7207889556884766, "policy_loss": -0.001022417795612185, "value_loss": 3.4723560214042664}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:39:47.188947+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985187.1889591, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1116820847863953, "elapsed_seconds": 7602.055505231023, "env_steps_per_second": 10523.469599104017, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:42:59.778578+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985379.7785916, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.111703482349177, "elapsed_seconds": 7602.132536457037, "env_steps_per_second": 10523.362966424142, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:42:59.855609+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985379.8556223, "update": 20000} diff --git a/non_social/seed_0000/attempt_01/status.json b/non_social/seed_0000/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..4891f24c7057abd7e3db73ab6c3be71757b3e9ed --- /dev/null +++ b/non_social/seed_0000/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.111703482349177, + "elapsed_seconds": 7602.132536457037, + "env_steps_per_second": 10523.362966424142, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0002434901476933704, + "chaser_grad_norm": 0.07690595369786024, + "chaser_new_fields": 83.0999984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.540000915527344, + "chaser_subspace_norm": 75.278076171875, + "collisions_per_episode": 1.25, + "entropy": 0.03620169032365084, + "explorer_grad_norm": 0.2844117097556591, + "explorer_new_fields": 75.5250015258789, + "explorer_partner_vision": 0.0, + "explorer_return": 62.76250076293945, + "explorer_subspace_norm": 28.751102447509766, + "final_distance": 5.306502342224121, + "max_param_abs": 1.7384926080703735, + "policy_loss": -0.0007488939012390006, + "value_loss": 3.4187601804733276 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:42:59.855609+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985379.8556223, + "update": 20000 +} diff --git a/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6d2d4df5d64d67383e5accae0bfa178c44fc6380 --- /dev/null +++ b/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a6c8227e6a0fa824f91e1ee406af15f03ca2f65ee00d67445626986d7698c1c3 +size 11460368 diff --git a/non_social/seed_0001/attempt_01/config.json b/non_social/seed_0001/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8fcbc992dc4dfc9488ed5567e017e3586aeb19a9 --- /dev/null +++ b/non_social/seed_0001/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", + "seed": 1, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0001/attempt_01/metrics.jsonl b/non_social/seed_0001/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..b4d2a79f269986a52829f5222fe7c1dbfb74c6f4 --- /dev/null +++ b/non_social/seed_0001/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006646091572474689, "elapsed_seconds": 2.392592966090888, "env_steps_per_second": 1671.8263644047056, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.291518535792129, "eta_seconds": 47849.466728851665, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -5.105864413525174e-06, "chaser_grad_norm": 0.35422321408987045, "chaser_new_fields": 34.47500228881836, "chaser_partner_vision": 0.0, "chaser_return": -3.140000343322754, "chaser_subspace_norm": 0.0020734171848744154, "collisions_per_episode": 1.100000023841858, "entropy": 2.7725194096565247, "explorer_grad_norm": 0.10158612206578255, "explorer_new_fields": 36.20000076293945, "explorer_partner_vision": 0.0, "explorer_return": 4.037499904632568, "explorer_subspace_norm": 0.002020237036049366, "final_distance": 5.169232368469238, "max_param_abs": 0.1542729288339615, "policy_loss": -0.0005274267090716878, "value_loss": 2.9001960158348083}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:20.165036+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977780.1650496, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05299670888110995, "elapsed_seconds": 190.78815197199583, "env_steps_per_second": 10482.83124149954, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.066871646363288, "eta_seconds": 7440.737926907837, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00024995170555008883, "chaser_grad_norm": 0.2540866546332836, "chaser_new_fields": 76.875, "chaser_partner_vision": 0.0, "chaser_return": 5.330000877380371, "chaser_subspace_norm": 17.4329891204834, "collisions_per_episode": 1.0500000715255737, "entropy": 0.39347945898771286, "explorer_grad_norm": 0.6242107301950455, "explorer_new_fields": 66.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 48.625, "explorer_subspace_norm": 9.651309967041016, "final_distance": 5.658512592315674, "max_param_abs": 0.3182811737060547, "policy_loss": -0.0018509065173546446, "value_loss": 4.656160950660706}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:28.560596+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977968.5606086, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10618631221250527, "elapsed_seconds": 382.270723965019, "env_steps_per_second": 10463.788486104508, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0175399320376, "eta_seconds": 7263.143755335361, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00034994851500846913, "chaser_grad_norm": 0.19566128961741924, "chaser_new_fields": 79.25, "chaser_partner_vision": 0.0, "chaser_return": 5.7950005531311035, "chaser_subspace_norm": 29.966073989868164, "collisions_per_episode": 1.0750000476837158, "entropy": 0.26417846977710724, "explorer_grad_norm": 0.7438876032829285, "explorer_new_fields": 66.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 50.275001525878906, "explorer_subspace_norm": 10.79336166381836, "final_distance": 5.03670072555542, "max_param_abs": 0.3332476317882538, "policy_loss": -0.0016920956687584976, "value_loss": 4.247400283813477}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:40.043166+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978160.0431795, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16047670955973445, "elapsed_seconds": 577.716154415044, "env_steps_per_second": 10385.72308242824, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9792127512367248, "eta_seconds": 7125.165904452209, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0008375835176397395, "chaser_grad_norm": 0.18825266137719154, "chaser_new_fields": 81.625, "chaser_partner_vision": 0.0, "chaser_return": 6.27000093460083, "chaser_subspace_norm": 40.897972106933594, "collisions_per_episode": 1.25, "entropy": 0.256906658411026, "explorer_grad_norm": 0.5816270112991333, "explorer_new_fields": 65.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 47.72500228881836, "explorer_subspace_norm": 11.335468292236328, "final_distance": 4.564890384674072, "max_param_abs": 0.4130561649799347, "policy_loss": -0.0011627619122265287, "value_loss": 4.187922835350037}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:55.488596+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978355.488607, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21357180113195984, "elapsed_seconds": 768.8584840750555, "env_steps_per_second": 10405.035732452223, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9221462101876385, "eta_seconds": 6919.726356675499, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0005999028207475943, "chaser_grad_norm": 0.20418540574610233, "chaser_new_fields": 81.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.260001182556152, "chaser_subspace_norm": 47.05528259277344, "collisions_per_episode": 1.3000000715255737, "entropy": 0.20487841591238976, "explorer_grad_norm": 0.7134600207209587, "explorer_new_fields": 67.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 50.95000076293945, "explorer_subspace_norm": 11.689831733703613, "final_distance": 5.1953253746032715, "max_param_abs": 0.46658068895339966, "policy_loss": -0.0004994259075061791, "value_loss": 3.8595205545425415}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:06.630927+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978546.6309407, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26611383178057924, "elapsed_seconds": 958.0097944100853, "env_steps_per_second": 10438.306641904126, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8627968224640548, "eta_seconds": 6706.068560870597, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002379925559409246, "chaser_grad_norm": 0.15856351889669895, "chaser_new_fields": 81.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.160001277923584, "chaser_subspace_norm": 47.54892349243164, "collisions_per_episode": 1.3000000715255737, "entropy": 0.19381681457161903, "explorer_grad_norm": 1.4128717184066772, "explorer_new_fields": 68.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 52.45000076293945, "explorer_subspace_norm": 14.399904251098633, "final_distance": 4.250465393066406, "max_param_abs": 0.4694162905216217, "policy_loss": -0.0007240931175118703, "value_loss": 4.089799165725708}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:15.782237+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978735.7822506, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.31946793677974106, "elapsed_seconds": 1150.0845724070678, "env_steps_per_second": 10434.01527844567, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8103183084185326, "eta_seconds": 6517.145910306717, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005075695641181543, "chaser_grad_norm": 0.0860081585124135, "chaser_new_fields": 82.875, "chaser_partner_vision": 0.0, "chaser_return": 6.490001201629639, "chaser_subspace_norm": 47.72903060913086, "collisions_per_episode": 1.8000000715255737, "entropy": 0.1328066848218441, "explorer_grad_norm": 0.5817417502403259, "explorer_new_fields": 71.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 56.38750076293945, "explorer_subspace_norm": 14.815073013305664, "final_distance": 4.913993835449219, "max_param_abs": 0.5209510326385498, "policy_loss": -0.00031853549470550746, "value_loss": 3.6523645520210266}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:27.857015+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978927.8570433, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.37271470920835986, "elapsed_seconds": 1341.7729531500954, "env_steps_per_second": 10433.956033419845, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7570836291251248, "eta_seconds": 6325.501064850449, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00013817141942573663, "chaser_grad_norm": 0.23563339933753014, "chaser_new_fields": 83.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.530001163482666, "chaser_subspace_norm": 46.80853271484375, "collisions_per_episode": 1.375, "entropy": 0.10780291259288788, "explorer_grad_norm": 0.5218307226896286, "explorer_new_fields": 71.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 56.04999923706055, "explorer_subspace_norm": 16.33576202392578, "final_distance": 4.8094482421875, "max_param_abs": 0.637711226940155, "policy_loss": -0.0008312481108596614, "value_loss": 3.6552406549453735}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:39.545395+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979119.5454068, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.42564808120805536, "elapsed_seconds": 1532.3330923489993, "env_steps_per_second": 10441.593984942727, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7025923248322214, "eta_seconds": 6129.332369395997, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": -5.299170846001294e-05, "chaser_grad_norm": 0.1636909916996956, "chaser_new_fields": 81.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.30000114440918, "chaser_subspace_norm": 55.12071228027344, "collisions_per_episode": 1.725000023841858, "entropy": 0.12000446207821369, "explorer_grad_norm": 1.3998910635709763, "explorer_new_fields": 71.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.28750228881836, "explorer_subspace_norm": 16.26192283630371, "final_distance": 4.884814739227295, "max_param_abs": 0.7220480442047119, "policy_loss": -0.0007309603997214253, "value_loss": 3.6593281626701355}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:50.105535+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979310.105551, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.47814068301862184, "elapsed_seconds": 1721.3064588670386, "env_steps_per_second": 10457.173333241062, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6469290192863644, "eta_seconds": 5928.944469430911, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 7.421613951296563e-05, "chaser_grad_norm": 0.23856427893042564, "chaser_new_fields": 81.125, "chaser_partner_vision": 0.0, "chaser_return": 6.140000820159912, "chaser_subspace_norm": 55.25377655029297, "collisions_per_episode": 1.475000023841858, "entropy": 0.1395217962563038, "explorer_grad_norm": 1.1217033490538597, "explorer_new_fields": 68.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 52.26250076293945, "explorer_subspace_norm": 17.047136306762695, "final_distance": 4.939172744750977, "max_param_abs": 0.8441739678382874, "policy_loss": -0.00040146764449322525, "value_loss": 3.8861085176467896}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:04:59.078901+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979499.078913, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5305872613688957, "elapsed_seconds": 1910.1141409280244, "env_steps_per_second": 10470.57847039604, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5917617841066871, "eta_seconds": 5730.342422784073, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0001957275774418577, "chaser_grad_norm": 0.06718951277434826, "chaser_new_fields": 79.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.835000514984131, "chaser_subspace_norm": 61.439979553222656, "collisions_per_episode": 1.1750000715255737, "entropy": 0.09594621136784554, "explorer_grad_norm": 0.5464240312576294, "explorer_new_fields": 68.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 53.01250076293945, "explorer_subspace_norm": 17.778528213500977, "final_distance": 5.9304118156433105, "max_param_abs": 0.9190318584442139, "policy_loss": -0.0004385370368851227, "value_loss": 3.810237228870392}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:07.886584+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979687.8865972, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5838003536613865, "elapsed_seconds": 2101.6812731809914, "env_steps_per_second": 10467.809881896119, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.53911002328911, "eta_seconds": 5540.796083840796, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0007284536260159166, "chaser_grad_norm": 0.19790029153227806, "chaser_new_fields": 79.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.900001049041748, "chaser_subspace_norm": 60.71714782714844, "collisions_per_episode": 1.6750000715255737, "entropy": 0.09285321645438671, "explorer_grad_norm": 1.190895400941372, "explorer_new_fields": 71.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 56.95000076293945, "explorer_subspace_norm": 19.046768188476562, "final_distance": 4.46820592880249, "max_param_abs": 0.9952329993247986, "policy_loss": -0.0007394702993295255, "value_loss": 3.7340222597122192}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:19.453715+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979879.4537265, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.637212698103071, "elapsed_seconds": 2293.9657131710555, "env_steps_per_second": 10462.231349928803, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4868296289071656, "eta_seconds": 5352.586664065796, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00012293299168864014, "chaser_grad_norm": 0.1078545544296503, "chaser_new_fields": 79.125, "chaser_partner_vision": 0.0, "chaser_return": 5.755000591278076, "chaser_subspace_norm": 68.20173645019531, "collisions_per_episode": 1.25, "entropy": 0.1013022381812334, "explorer_grad_norm": 0.5451182201504707, "explorer_new_fields": 70.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 55.41250228881836, "explorer_subspace_norm": 17.95201301574707, "final_distance": 5.370412349700928, "max_param_abs": 1.0787969827651978, "policy_loss": -0.0004317292224147451, "value_loss": 3.706162452697754}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:31.738157+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980071.738171, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6894948651758346, "elapsed_seconds": 2482.1815146330046, "env_steps_per_second": 10474.657009056065, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4320277969036563, "eta_seconds": 5155.300068853163, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0002024353854959693, "chaser_grad_norm": 0.2383432388305664, "chaser_new_fields": 76.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.325000762939453, "chaser_subspace_norm": 65.47434997558594, "collisions_per_episode": 1.399999976158142, "entropy": 0.09009139984846115, "explorer_grad_norm": 2.05173921585083, "explorer_new_fields": 74.0, "explorer_partner_vision": 0.0, "explorer_return": 60.400001525878906, "explorer_subspace_norm": 18.082992553710938, "final_distance": 5.4761481285095215, "max_param_abs": 1.11553156375885, "policy_loss": -0.0003861995520466621, "value_loss": 3.5980712175369263}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:39.953957+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980259.9539683, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.741385722357243, "elapsed_seconds": 2668.9886004860746, "env_steps_per_second": 10490.865339365128, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3768591986634513, "eta_seconds": 4956.693115188425, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00014662955988614668, "chaser_grad_norm": 0.13823741115629673, "chaser_new_fields": 76.375, "chaser_partner_vision": 0.0, "chaser_return": 5.185000896453857, "chaser_subspace_norm": 81.42784118652344, "collisions_per_episode": 1.1750000715255737, "entropy": 0.08847093023359776, "explorer_grad_norm": 0.4604616239666939, "explorer_new_fields": 75.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 62.61249923706055, "explorer_subspace_norm": 21.18718147277832, "final_distance": 6.204265117645264, "max_param_abs": 1.158452033996582, "policy_loss": -0.0010841385633639788, "value_loss": 4.00962769985199}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:20:46.761043+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980446.7610588, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7941792751552809, "elapsed_seconds": 2859.045390559011, "env_steps_per_second": 10493.012842350952, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3236321252588015, "eta_seconds": 4765.075650931685, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 2.783902959220541e-05, "chaser_grad_norm": 0.07786290813237429, "chaser_new_fields": 83.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.655002117156982, "chaser_subspace_norm": 94.4039306640625, "collisions_per_episode": 1.024999976158142, "entropy": 0.09832780249416828, "explorer_grad_norm": 1.3362184911966324, "explorer_new_fields": 75.625, "explorer_partner_vision": 0.0, "explorer_return": 63.0625, "explorer_subspace_norm": 21.0552921295166, "final_distance": 5.50551176071167, "max_param_abs": 1.0359666347503662, "policy_loss": -0.00043181558649152407, "value_loss": 3.7632511854171753}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:23:56.817832+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980636.8178422, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8470003187508296, "elapsed_seconds": 3049.2011475029867, "env_steps_per_second": 10494.551999695079, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2705004781262446, "eta_seconds": 4573.80172125448, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0004685354021869981, "chaser_grad_norm": 0.15229122154414654, "chaser_new_fields": 82.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.4500017166137695, "chaser_subspace_norm": 96.91693878173828, "collisions_per_episode": 1.0500000715255737, "entropy": 0.06496398337185383, "explorer_grad_norm": 1.082899585366249, "explorer_new_fields": 76.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 63.70000076293945, "explorer_subspace_norm": 23.27690887451172, "final_distance": 5.564136505126953, "max_param_abs": 0.9712896347045898, "policy_loss": -0.0011512674705596027, "value_loss": 3.5662338733673096}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:06.973591+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980826.9736054, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8991850248044486, "elapsed_seconds": 3237.066089296015, "env_steps_per_second": 10503.3382273002, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2165444453236658, "eta_seconds": 4379.560003165197, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0002205174978130442, "chaser_grad_norm": 0.08294959180057049, "chaser_new_fields": 81.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.190001010894775, "chaser_subspace_norm": 101.9328384399414, "collisions_per_episode": 1.0, "entropy": 0.045907335355877876, "explorer_grad_norm": 0.5273615941405296, "explorer_new_fields": 77.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 65.1624984741211, "explorer_subspace_norm": 22.970808029174805, "final_distance": 5.270810127258301, "max_param_abs": 0.974787712097168, "policy_loss": -0.0005561281775072757, "value_loss": 3.466669023036957}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:14.838531+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981014.8385422, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9520237867916714, "elapsed_seconds": 3427.285632450017, "env_steps_per_second": 10503.93922792632, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.163584628300932, "eta_seconds": 4188.9046618833545, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 5.521597716218185e-05, "chaser_grad_norm": 0.13706197775900364, "chaser_new_fields": 83.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.675000190734863, "chaser_subspace_norm": 95.42662811279297, "collisions_per_episode": 0.8500000238418579, "entropy": 0.04296411946415901, "explorer_grad_norm": 0.41673433035612106, "explorer_new_fields": 77.375, "explorer_partner_vision": 0.0, "explorer_return": 65.76250457763672, "explorer_subspace_norm": 25.36161994934082, "final_distance": 5.151485443115234, "max_param_abs": 0.9597460031509399, "policy_loss": -0.0002633670809331079, "value_loss": 3.5941293239593506}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:25.058075+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981205.058089, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0056366006861208, "elapsed_seconds": 3620.291762470035, "env_steps_per_second": 10496.391587531483, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1114930849688704, "eta_seconds": 4001.3751058879334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00012517708195831923, "chaser_grad_norm": 0.2932741940021515, "chaser_new_fields": 72.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.380000591278076, "chaser_subspace_norm": 89.20765686035156, "collisions_per_episode": 1.0, "entropy": 0.0684291161596775, "explorer_grad_norm": 1.9711225181818008, "explorer_new_fields": 73.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 59.462501525878906, "explorer_subspace_norm": 24.477724075317383, "final_distance": 5.2902936935424805, "max_param_abs": 0.9854035973548889, "policy_loss": -0.0001851809129824744, "value_loss": 3.8698012232780457}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:38.064206+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981398.0642197, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0596000968719534, "elapsed_seconds": 3814.5603487390326, "env_steps_per_second": 10486.13636777897, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0596000968719534, "eta_seconds": 3814.560348739032, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00038964367026812774, "chaser_grad_norm": 0.09080993384122849, "chaser_new_fields": 80.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.085000991821289, "chaser_subspace_norm": 79.93997192382812, "collisions_per_episode": 1.024999976158142, "entropy": 0.05338955298066139, "explorer_grad_norm": 0.8891770392656326, "explorer_new_fields": 75.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 62.3125, "explorer_subspace_norm": 26.096323013305664, "final_distance": 5.22806978225708, "max_param_abs": 0.9808329343795776, "policy_loss": -0.0008176689683470073, "value_loss": 3.6155094504356384}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:39:52.332791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981592.3328056, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.113099402428359, "elapsed_seconds": 4007.157848742092, "env_steps_per_second": 10481.244209829278, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.00708993553042, "eta_seconds": 3625.523767909512, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00045353863636776737, "chaser_grad_norm": 0.06808555126190186, "chaser_new_fields": 78.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.690001010894775, "chaser_subspace_norm": 90.07524871826172, "collisions_per_episode": 1.4250000715255737, "entropy": 0.056026156060397625, "explorer_grad_norm": 0.768612265586853, "explorer_new_fields": 73.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 58.97500228881836, "explorer_subspace_norm": 24.316312789916992, "final_distance": 4.746428966522217, "max_param_abs": 0.9814541339874268, "policy_loss": -0.0005006787593009321, "value_loss": 3.4611494541168213}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:04.930293+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981784.9303074, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.166562008252238, "elapsed_seconds": 4199.623229708057, "env_steps_per_second": 10477.130350347816, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9544598249336493, "eta_seconds": 3436.0553697611376, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0002189835984244791, "chaser_grad_norm": 0.14183006435632706, "chaser_new_fields": 80.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.065001010894775, "chaser_subspace_norm": 91.1957015991211, "collisions_per_episode": 2.2250001430511475, "entropy": 0.09379604086279869, "explorer_grad_norm": 0.3437352776527405, "explorer_new_fields": 73.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 58.97500228881836, "explorer_subspace_norm": 25.498416900634766, "final_distance": 4.4637041091918945, "max_param_abs": 0.9716539978981018, "policy_loss": -0.0009403482038745636, "value_loss": 3.6964353919029236}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:17.395673+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981977.3956897, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2201823734194557, "elapsed_seconds": 4392.656544310041, "env_steps_per_second": 10472.02291733584, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9018739281795977, "eta_seconds": 3246.7461414465515, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0007552657056391077, "chaser_grad_norm": 0.058179725892841816, "chaser_new_fields": 78.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.590001106262207, "chaser_subspace_norm": 84.99630737304688, "collisions_per_episode": 1.399999976158142, "entropy": 0.07308833859860897, "explorer_grad_norm": 1.2959906607866287, "explorer_new_fields": 77.5, "explorer_partner_vision": 0.0, "explorer_return": 65.57500457763672, "explorer_subspace_norm": 25.410776138305664, "final_distance": 5.0258612632751465, "max_param_abs": 1.0161798000335693, "policy_loss": -0.0010617657257228075, "value_loss": 3.5079697370529175}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:30.428988+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982170.4290025, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2742603790403033, "elapsed_seconds": 4587.337364545092, "env_steps_per_second": 10463.586212556653, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8495069193602024, "eta_seconds": 3058.2249096967284, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00021119906125931925, "chaser_grad_norm": 0.13109773769974709, "chaser_new_fields": 76.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.325000762939453, "chaser_subspace_norm": 82.43509674072266, "collisions_per_episode": 1.600000023841858, "entropy": 0.07194264605641365, "explorer_grad_norm": 1.371770866215229, "explorer_new_fields": 70.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 54.36249923706055, "explorer_subspace_norm": 23.883390426635742, "final_distance": 5.147349834442139, "max_param_abs": 1.032204270362854, "policy_loss": -0.0007983786445109331, "value_loss": 3.6320714354515076}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:45.109808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982365.109821, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.326985051559459, "elapsed_seconds": 4777.146185614052, "env_steps_per_second": 10466.49988450647, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7961910309356753, "eta_seconds": 2866.2877113684312, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 9.733648125020977e-05, "chaser_grad_norm": 0.21115468814969063, "chaser_new_fields": 76.125, "chaser_partner_vision": 0.0, "chaser_return": 5.125000953674316, "chaser_subspace_norm": 90.5826187133789, "collisions_per_episode": 1.899999976158142, "entropy": 0.07294890843331814, "explorer_grad_norm": 0.7679236233234406, "explorer_new_fields": 70.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 55.60000228881836, "explorer_subspace_norm": 23.094196319580078, "final_distance": 4.849133491516113, "max_param_abs": 1.0736829042434692, "policy_loss": -0.0006347609012173905, "value_loss": 3.693317234516144}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:54.918628+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982554.9186392, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3789171851091668, "elapsed_seconds": 4964.101866393001, "env_steps_per_second": 10475.208083871184, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7424938689049361, "eta_seconds": 2672.97792805777, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0007749467855087122, "chaser_grad_norm": 0.05683001130819321, "chaser_new_fields": 77.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.455001354217529, "chaser_subspace_norm": 94.65978240966797, "collisions_per_episode": 1.149999976158142, "entropy": 0.0547284884378314, "explorer_grad_norm": 0.8309146612882614, "explorer_new_fields": 74.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 61.1875, "explorer_subspace_norm": 24.733530044555664, "final_distance": 5.075489044189453, "max_param_abs": 1.0736836194992065, "policy_loss": -0.0010926405941162898, "value_loss": 3.685732841491699}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:01.874310+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982741.8743246, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.432759276984725, "elapsed_seconds": 5157.9333971450105, "env_steps_per_second": 10469.309283809242, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6898470592889417, "eta_seconds": 2483.4494134401903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0005101998941146879, "chaser_grad_norm": 0.09003976918756962, "chaser_new_fields": 77.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.4250006675720215, "chaser_subspace_norm": 89.87812042236328, "collisions_per_episode": 1.7000000476837158, "entropy": 0.11065679043531418, "explorer_grad_norm": 0.5014339312911034, "explorer_new_fields": 71.25, "explorer_partner_vision": 0.0, "explorer_return": 56.275001525878906, "explorer_subspace_norm": 21.806711196899414, "final_distance": 5.196125030517578, "max_param_abs": 1.0736504793167114, "policy_loss": -0.0009785575692120574, "value_loss": 3.7667288780212402}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:15.705839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982935.7058506, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4857316654825181, "elapsed_seconds": 5348.633995737066, "env_steps_per_second": 10469.962993286279, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6367421423496507, "eta_seconds": 2292.2717124587425, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0002474860123922312, "chaser_grad_norm": 0.05047455057501793, "chaser_new_fields": 77.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.3600006103515625, "chaser_subspace_norm": 87.39171600341797, "collisions_per_episode": 2.0250000953674316, "entropy": 0.08761332742869854, "explorer_grad_norm": 1.0886653810739517, "explorer_new_fields": 71.5, "explorer_partner_vision": 0.0, "explorer_return": 56.57500076293945, "explorer_subspace_norm": 22.313016891479492, "final_distance": 4.741565704345703, "max_param_abs": 1.0863597393035889, "policy_loss": -0.0005897007421591383, "value_loss": 3.664694607257843}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:26.406439+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983126.4064524, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5400618418230765, "elapsed_seconds": 5544.222630563076, "env_steps_per_second": 10461.340365422784, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5841613882777188, "eta_seconds": 2102.9809977997875, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0008129233688594084, "chaser_grad_norm": 0.08111646585166454, "chaser_new_fields": 76.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.185000896453857, "chaser_subspace_norm": 86.90794372558594, "collisions_per_episode": 1.9250000715255737, "entropy": 0.10414018481969833, "explorer_grad_norm": 0.7238684520125389, "explorer_new_fields": 68.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 51.85000228881836, "explorer_subspace_norm": 22.803674697875977, "final_distance": 5.295478820800781, "max_param_abs": 1.1287198066711426, "policy_loss": -0.0016683751654227308, "value_loss": 3.6987728476524353}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:41.995072+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983321.9950838, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5932511207977789, "elapsed_seconds": 5735.704034872004, "env_steps_per_second": 10460.790800085093, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5310837069325929, "eta_seconds": 1911.9013449573345, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0002415505494585135, "chaser_grad_norm": 0.061741966754198074, "chaser_new_fields": 77.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.490001201629639, "chaser_subspace_norm": 90.45314025878906, "collisions_per_episode": 1.774999976158142, "entropy": 0.06925754249095917, "explorer_grad_norm": 0.8019117414951324, "explorer_new_fields": 72.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 58.60000228881836, "explorer_subspace_norm": 25.05002784729004, "final_distance": 5.04298210144043, "max_param_abs": 1.1126190423965454, "policy_loss": -0.000777919802583682, "value_loss": 3.715272307395935}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:53.476478+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983513.476493, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6452307465797316, "elapsed_seconds": 5922.830687687034, "env_steps_per_second": 10467.967644067849, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4776476361037931, "eta_seconds": 1719.5314899736552, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0014523894221247918, "chaser_grad_norm": 0.07097857166081667, "chaser_new_fields": 79.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.785001277923584, "chaser_subspace_norm": 94.98880767822266, "collisions_per_episode": 1.350000023841858, "entropy": 0.07164299860596657, "explorer_grad_norm": 0.6239117234945297, "explorer_new_fields": 73.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 60.28750228881836, "explorer_subspace_norm": 25.258012771606445, "final_distance": 5.408947467803955, "max_param_abs": 1.1395074129104614, "policy_loss": -0.0011831774853590105, "value_loss": 3.4303807616233826}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:15:00.603130+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983700.6031415, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6975080731550245, "elapsed_seconds": 6111.029063358088, "env_steps_per_second": 10472.867881409025, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4243770182887561, "eta_seconds": 1527.757265839522, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0002821423777086274, "chaser_grad_norm": 0.0475468123331666, "chaser_new_fields": 77.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.400001049041748, "chaser_subspace_norm": 101.9891128540039, "collisions_per_episode": 1.350000023841858, "entropy": 0.04990239441394806, "explorer_grad_norm": 0.5156930722296238, "explorer_new_fields": 71.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 57.51250076293945, "explorer_subspace_norm": 27.15218734741211, "final_distance": 5.17217493057251, "max_param_abs": 1.1514462232589722, "policy_loss": -0.000498238339438295, "value_loss": 3.62851220369339}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:08.801506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983888.8015206, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.74950597281917, "elapsed_seconds": 6298.221502149012, "env_steps_per_second": 10479.148752942427, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.37110732756770276, "eta_seconds": 1335.98637924373, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0001662679765182147, "chaser_grad_norm": 0.058657112531363964, "chaser_new_fields": 77.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.4850006103515625, "chaser_subspace_norm": 94.71659088134766, "collisions_per_episode": 1.9250000715255737, "entropy": 0.06443462520837784, "explorer_grad_norm": 0.46494172513484955, "explorer_new_fields": 69.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 53.275001525878906, "explorer_subspace_norm": 27.593393325805664, "final_distance": 5.075628757476807, "max_param_abs": 1.1629002094268799, "policy_loss": -0.0009729938061013854, "value_loss": 3.6678024530410767}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:15.993944+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984075.9939582, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8016514310583525, "elapsed_seconds": 6485.945151810069, "env_steps_per_second": 10484.208300932496, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3179384878338269, "eta_seconds": 1144.5785562017768, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00023642260933431913, "chaser_grad_norm": 0.12356526777148247, "chaser_new_fields": 77.875, "chaser_partner_vision": 0.0, "chaser_return": 5.445001125335693, "chaser_subspace_norm": 79.68389129638672, "collisions_per_episode": 1.875, "entropy": 0.08755725808441639, "explorer_grad_norm": 0.9381912797689438, "explorer_new_fields": 72.5, "explorer_partner_vision": 0.0, "explorer_return": 58.0, "explorer_subspace_norm": 26.78290557861328, "final_distance": 5.436651706695557, "max_param_abs": 1.267048954963684, "policy_loss": -0.000974916765464684, "value_loss": 3.7329800724983215}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:23.717595+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984263.71761, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.854691653937787, "elapsed_seconds": 6676.889954176033, "env_steps_per_second": 10483.922976178272, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.264955950562541, "eta_seconds": 953.8414220251476, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 1.806048938457483e-07, "chaser_grad_norm": 0.050238268449902534, "chaser_new_fields": 80.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.945001125335693, "chaser_subspace_norm": 91.65946197509766, "collisions_per_episode": 1.8250000476837158, "entropy": 0.03836974501609802, "explorer_grad_norm": 0.7704875618219376, "explorer_new_fields": 74.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 60.32500076293945, "explorer_subspace_norm": 26.297279357910156, "final_distance": 5.851404666900635, "max_param_abs": 1.2902907133102417, "policy_loss": -0.00034309800324283657, "value_loss": 3.4310280084609985}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:34.662396+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984454.6624086, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9075580443783353, "elapsed_seconds": 6867.208959762007, "env_steps_per_second": 10484.608874126245, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21195089381981505, "eta_seconds": 763.0232177513342, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00015660543499332324, "chaser_grad_norm": 0.2612560838460922, "chaser_new_fields": 78.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.575000762939453, "chaser_subspace_norm": 96.18538665771484, "collisions_per_episode": 1.5750000476837158, "entropy": 0.058209532871842384, "explorer_grad_norm": 0.9300043731927872, "explorer_new_fields": 75.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 62.23749923706055, "explorer_subspace_norm": 26.099992752075195, "final_distance": 5.504971981048584, "max_param_abs": 1.2806016206741333, "policy_loss": -0.0008122388410924941, "value_loss": 3.572008788585663}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:30:44.981404+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984644.9814177, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9594223218633366, "elapsed_seconds": 7053.920358708012, "env_steps_per_second": 10490.620284456083, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15887208015108134, "eta_seconds": 571.9394885438928, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00017630407430990758, "chaser_grad_norm": 0.1251893099397421, "chaser_new_fields": 79.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.695001125335693, "chaser_subspace_norm": 95.90496063232422, "collisions_per_episode": 1.475000023841858, "entropy": 0.10102851688861847, "explorer_grad_norm": 0.8435127064585686, "explorer_new_fields": 68.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 51.70000076293945, "explorer_subspace_norm": 26.99217987060547, "final_distance": 4.385324954986572, "max_param_abs": 1.3020002841949463, "policy_loss": -0.0010951019522043026, "value_loss": 3.8845006227493286}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:33:51.692801+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984831.6928124, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0117075500569706, "elapsed_seconds": 7242.147180205095, "env_steps_per_second": 10494.125306888296, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10587934473984056, "eta_seconds": 381.16564106342605, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00022960279347063314, "chaser_grad_norm": 0.12079357635229826, "chaser_new_fields": 80.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.040000915527344, "chaser_subspace_norm": 102.02371215820312, "collisions_per_episode": 1.5, "entropy": 0.07814025320112705, "explorer_grad_norm": 1.5280047059059143, "explorer_new_fields": 72.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 58.29999923706055, "explorer_subspace_norm": 24.30501365661621, "final_distance": 4.763765335083008, "max_param_abs": 1.2802588939666748, "policy_loss": -0.0002921351334101985, "value_loss": 3.4545443654060364}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:36:59.919623+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985019.919637, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.064875669445286, "elapsed_seconds": 7433.552410003031, "env_steps_per_second": 10492.964291882648, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05294552998577656, "eta_seconds": 190.60390794879564, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0006159323773005099, "chaser_grad_norm": 0.057113331742584705, "chaser_new_fields": 81.625, "chaser_partner_vision": 0.0, "chaser_return": 6.265000820159912, "chaser_subspace_norm": 100.68077850341797, "collisions_per_episode": 0.9000000357627869, "entropy": 0.07017142325639725, "explorer_grad_norm": 1.1960067749023438, "explorer_new_fields": 71.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.10000228881836, "explorer_subspace_norm": 27.36815071105957, "final_distance": 4.832053184509277, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0004417272417356344, "value_loss": 3.56430321931839}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:11.324852+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985211.324863, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1162682046338968, "elapsed_seconds": 7618.565536682028, "env_steps_per_second": 10500.664411799613, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:16.337978+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985396.3379903, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.116271049076135, "elapsed_seconds": 7618.575776674086, "env_steps_per_second": 10500.650298043534, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:16.348219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985396.3482296, "update": 20000} diff --git a/non_social/seed_0001/attempt_01/status.json b/non_social/seed_0001/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..bb6c20975c9f3cd50ebcea0ddfd45ff9a8b383e3 --- /dev/null +++ b/non_social/seed_0001/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.116271049076135, + "elapsed_seconds": 7618.575776674086, + "env_steps_per_second": 10500.650298043534, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0007710583850932262, + "chaser_grad_norm": 0.11076641827821732, + "chaser_new_fields": 83.25, + "chaser_partner_vision": 0.0, + "chaser_return": 6.595000743865967, + "chaser_subspace_norm": 107.82763671875, + "collisions_per_episode": 0.9750000238418579, + "entropy": 0.06922571919858456, + "explorer_grad_norm": 0.3988708183169365, + "explorer_new_fields": 73.5, + "explorer_partner_vision": 0.0, + "explorer_return": 59.76250076293945, + "explorer_subspace_norm": 27.074764251708984, + "final_distance": 4.443612575531006, + "max_param_abs": 1.3260666131973267, + "policy_loss": -0.0006433645908252572, + "value_loss": 3.4030256271362305 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:16.348219+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985396.3482296, + "update": 20000 +} diff --git a/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c75451e757144c9decff1d22f4c90ac5e7448b51 --- /dev/null +++ b/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e94060644c4de0c5467f9d15df5a5e4cfc6fbff18dcf358eb5cebe1dce94d3d7 +size 11460384 diff --git a/non_social/seed_0002/attempt_01/config.json b/non_social/seed_0002/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..0ee576445bddea2ef804b9ad457660f284ccf36e --- /dev/null +++ b/non_social/seed_0002/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", + "seed": 2, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0002/attempt_01/metrics.jsonl b/non_social/seed_0002/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..779ca88648061d05190bfcd779ae17de14a36639 --- /dev/null +++ b/non_social/seed_0002/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.00041805490165845385, "elapsed_seconds": 1.504997645970434, "env_steps_per_second": 2657.8114661573236, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 8.36067997826742, "eta_seconds": 30098.447921762712, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 4.5359129329895775e-05, "chaser_grad_norm": 0.3901529684662819, "chaser_new_fields": 33.75, "chaser_partner_vision": 0.0, "chaser_return": -3.2950005531311035, "chaser_subspace_norm": 0.0027342543471604586, "collisions_per_episode": 1.8000000715255737, "entropy": 2.7724889516830444, "explorer_grad_norm": 0.06409388780593872, "explorer_new_fields": 36.75, "explorer_partner_vision": 0.0, "explorer_return": 4.974999904632568, "explorer_subspace_norm": 0.0024781289976090193, "final_distance": 5.509538173675537, "max_param_abs": 0.15428684651851654, "policy_loss": -0.0005347863633655692, "value_loss": 3.1029520630836487}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.124633+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.1246529, "update": 1} +{"device": "cuda", "elapsed_hours": 0.052658536858313405, "elapsed_seconds": 189.57073268992826, "env_steps_per_second": 10550.151764572773, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.053682937474223, "eta_seconds": 7393.258574907202, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00034477247292752367, "chaser_grad_norm": 0.2059476338326931, "chaser_new_fields": 75.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.9850006103515625, "chaser_subspace_norm": 15.376718521118164, "collisions_per_episode": 2.25, "entropy": 0.36665698885917664, "explorer_grad_norm": 0.34435340762138367, "explorer_new_fields": 63.79999923706055, "explorer_partner_vision": 0.0, "explorer_return": 44.837501525878906, "explorer_subspace_norm": 12.487297058105469, "final_distance": 5.4530863761901855, "max_param_abs": 0.31553390622138977, "policy_loss": -0.0013678885430579157, "value_loss": 4.764330863952637}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:27.190366+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977967.1903799, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10514469126721249, "elapsed_seconds": 378.52088856196497, "env_steps_per_second": 10567.448510427948, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9977491340770372, "eta_seconds": 7191.896882677334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0007091937250672253, "chaser_grad_norm": 0.13627085834741592, "chaser_new_fields": 76.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.260000705718994, "chaser_subspace_norm": 30.3323917388916, "collisions_per_episode": 2.625, "entropy": 0.2446039579808712, "explorer_grad_norm": 0.5137593448162079, "explorer_new_fields": 65.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 47.6875, "explorer_subspace_norm": 13.4423189163208, "final_distance": 5.27152681350708, "max_param_abs": 0.3203819990158081, "policy_loss": -0.0023872482660323158, "value_loss": 4.376429200172424}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:36.140521+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978156.140534, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.1585148403063714, "elapsed_seconds": 570.653425102937, "env_steps_per_second": 10514.262661119914, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9550163637785805, "eta_seconds": 7038.05890960289, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00030876674687352335, "chaser_grad_norm": 0.13776557333767414, "chaser_new_fields": 74.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.755000591278076, "chaser_subspace_norm": 40.69841766357422, "collisions_per_episode": 2.575000047683716, "entropy": 0.1835484318435192, "explorer_grad_norm": 0.44421710819005966, "explorer_new_fields": 68.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 51.775001525878906, "explorer_subspace_norm": 15.046452522277832, "final_distance": 5.161291599273682, "max_param_abs": 0.351313978433609, "policy_loss": -0.001704655248673248, "value_loss": 4.042437434196472}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:48.273055+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978348.2730665, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21247627752192785, "elapsed_seconds": 764.9145990789402, "env_steps_per_second": 10458.683897042982, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9122864976973506, "eta_seconds": 6884.231391710462, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 3.8977193285383294e-05, "chaser_grad_norm": 0.13531122170388699, "chaser_new_fields": 71.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.15500020980835, "chaser_subspace_norm": 43.64929962158203, "collisions_per_episode": 3.25, "entropy": 0.2015102617442608, "explorer_grad_norm": 0.342316634953022, "explorer_new_fields": 69.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 53.5, "explorer_subspace_norm": 16.92076873779297, "final_distance": 5.106712341308594, "max_param_abs": 0.42098745703697205, "policy_loss": -0.0014348333568747762, "value_loss": 4.173529028892517}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:02.534233+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978542.534249, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26546482266803245, "elapsed_seconds": 955.6733616049169, "env_steps_per_second": 10463.826242060812, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8582537586762273, "eta_seconds": 6689.713531234418, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 4.6662839026045955e-05, "chaser_grad_norm": 0.2006438672542572, "chaser_new_fields": 75.375, "chaser_partner_vision": 0.0, "chaser_return": 4.935000896453857, "chaser_subspace_norm": 48.39346694946289, "collisions_per_episode": 2.450000047683716, "entropy": 0.1323208101093769, "explorer_grad_norm": 1.4133809953927994, "explorer_new_fields": 74.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 60.4375, "explorer_subspace_norm": 17.8142032623291, "final_distance": 5.531247615814209, "max_param_abs": 0.47428324818611145, "policy_loss": -0.0011846383656362036, "value_loss": 3.833590030670166}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:13.292994+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978733.2930071, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3190474914188639, "elapsed_seconds": 1148.57096910791, "env_steps_per_second": 10447.765373453889, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8079357847068955, "eta_seconds": 6508.568824944824, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": -6.051846308147146e-05, "chaser_grad_norm": 0.1595793403685093, "chaser_new_fields": 77.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.450000762939453, "chaser_subspace_norm": 51.27262496948242, "collisions_per_episode": 2.200000047683716, "entropy": 0.09537966363132, "explorer_grad_norm": 1.0771355703473091, "explorer_new_fields": 74.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.75, "explorer_subspace_norm": 18.244585037231445, "final_distance": 5.197914123535156, "max_param_abs": 0.5079510807991028, "policy_loss": -0.001016846292642981, "value_loss": 3.7896552085876465}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:26.190602+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978926.1906168, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3724662167769404, "elapsed_seconds": 1340.8783803969854, "env_steps_per_second": 10440.917091865638, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.755912164805576, "eta_seconds": 6321.283793300074, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0007547492599169203, "chaser_grad_norm": 0.08221679460257292, "chaser_new_fields": 75.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.000000953674316, "chaser_subspace_norm": 50.85700607299805, "collisions_per_episode": 2.4000000953674316, "entropy": 0.09722137078642845, "explorer_grad_norm": 0.44746997207403183, "explorer_new_fields": 71.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 55.97500228881836, "explorer_subspace_norm": 21.523387908935547, "final_distance": 5.273843288421631, "max_param_abs": 0.5621373653411865, "policy_loss": -0.0016603976006202004, "value_loss": 4.076526641845703}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:38.498011+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979118.4980233, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.42454720041417104, "elapsed_seconds": 1528.3699214910157, "env_steps_per_second": 10468.669773604972, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6981888016566842, "eta_seconds": 6113.479685964063, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0001353911080619885, "chaser_grad_norm": 0.15590168721973896, "chaser_new_fields": 76.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.040000915527344, "chaser_subspace_norm": 50.92699432373047, "collisions_per_episode": 3.200000047683716, "entropy": 0.10538040474057198, "explorer_grad_norm": 0.4895930662751198, "explorer_new_fields": 71.875, "explorer_partner_vision": 0.0, "explorer_return": 56.98749923706055, "explorer_subspace_norm": 22.22146987915039, "final_distance": 5.148885726928711, "max_param_abs": 0.586829423904419, "policy_loss": -0.0007573493281363142, "value_loss": 3.9318071603775024}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:45.989553+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979305.9895663, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.47820473831748433, "elapsed_seconds": 1721.5370579429436, "env_steps_per_second": 10455.772599811538, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6471496542046684, "eta_seconds": 5929.7387551368065, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00018599914956105024, "chaser_grad_norm": 0.3318144716322422, "chaser_new_fields": 74.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.830000877380371, "chaser_subspace_norm": 57.51749038696289, "collisions_per_episode": 2.0250000953674316, "entropy": 0.0917278565466404, "explorer_grad_norm": 0.7204478234052658, "explorer_new_fields": 74.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 61.150001525878906, "explorer_subspace_norm": 23.193132400512695, "final_distance": 4.897786617279053, "max_param_abs": 0.586829423904419, "policy_loss": -0.0010835970030171893, "value_loss": 3.991298019886017}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:04:59.156688+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979499.1566987, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5308737403380412, "elapsed_seconds": 1911.1454652169487, "env_steps_per_second": 10464.928161671696, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.592621221014124, "eta_seconds": 5733.436395650846, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00018661705734013267, "chaser_grad_norm": 0.1579245738685131, "chaser_new_fields": 78.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.465001106262207, "chaser_subspace_norm": 54.17221450805664, "collisions_per_episode": 3.0250000953674316, "entropy": 0.09232253767549992, "explorer_grad_norm": 0.8321195915341377, "explorer_new_fields": 72.25, "explorer_partner_vision": 0.0, "explorer_return": 57.47500228881836, "explorer_subspace_norm": 25.65125274658203, "final_distance": 4.809153079986572, "max_param_abs": 0.6354391574859619, "policy_loss": -0.0011576689546899388, "value_loss": 3.7713403701782227}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:08.765096+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979688.7651098, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5827764931577581, "elapsed_seconds": 2097.9953753679292, "env_steps_per_second": 10486.200426510388, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.536410754688635, "eta_seconds": 5531.078716879086, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00023536643245702926, "chaser_grad_norm": 0.15026073344051838, "chaser_new_fields": 77.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.405000686645508, "chaser_subspace_norm": 57.10222625732422, "collisions_per_episode": 2.075000047683716, "entropy": 0.0711546279489994, "explorer_grad_norm": 0.7971340417861938, "explorer_new_fields": 74.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 60.587501525878906, "explorer_subspace_norm": 26.435279846191406, "final_distance": 4.737027645111084, "max_param_abs": 0.6828851103782654, "policy_loss": -0.0007055611645228055, "value_loss": 3.8577597737312317}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:15.615006+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979875.6150181, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6346422604947131, "elapsed_seconds": 2284.712137780967, "env_steps_per_second": 10504.605636362603, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4808319411543307, "eta_seconds": 5330.99498815559, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 4.201918466506527e-07, "chaser_grad_norm": 0.16751500964164734, "chaser_new_fields": 77.875, "chaser_partner_vision": 0.0, "chaser_return": 5.440001010894775, "chaser_subspace_norm": 59.215232849121094, "collisions_per_episode": 2.375, "entropy": 0.08229825459420681, "explorer_grad_norm": 0.45556944608688354, "explorer_new_fields": 73.625, "explorer_partner_vision": 0.0, "explorer_return": 59.875, "explorer_subspace_norm": 28.377948760986328, "final_distance": 4.696279048919678, "max_param_abs": 0.6842984557151794, "policy_loss": -0.00042481566849517094, "value_loss": 4.050582528114319}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:22.331769+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980062.3317828, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6872818397705366, "elapsed_seconds": 2474.2146231739316, "env_steps_per_second": 10508.385067519772, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4274315133695759, "eta_seconds": 5138.753448130473, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0004071449551360101, "chaser_grad_norm": 0.1661434005945921, "chaser_new_fields": 79.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.8650007247924805, "chaser_subspace_norm": 65.4849624633789, "collisions_per_episode": 2.325000047683716, "entropy": 0.056140294298529625, "explorer_grad_norm": 0.428345762193203, "explorer_new_fields": 79.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 68.01250457763672, "explorer_subspace_norm": 30.894672393798828, "final_distance": 5.670973777770996, "max_param_abs": 0.728032112121582, "policy_loss": -0.00036883605515924955, "value_loss": 3.7681143283843994}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:31.834253+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980251.8342643, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7391429485713726, "elapsed_seconds": 2660.9146148569416, "env_steps_per_second": 10522.697663301518, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.372694047346835, "eta_seconds": 4941.698570448606, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0005906517320033236, "chaser_grad_norm": 0.17412240616977215, "chaser_new_fields": 72.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.460000514984131, "chaser_subspace_norm": 71.85272216796875, "collisions_per_episode": 2.625, "entropy": 0.049920396879315376, "explorer_grad_norm": 0.4346041679382324, "explorer_new_fields": 80.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 69.36250305175781, "explorer_subspace_norm": 32.59563446044922, "final_distance": 5.265743732452393, "max_param_abs": 0.765934944152832, "policy_loss": -0.0009554296231470971, "value_loss": 3.815235674381256}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:20:38.534247+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980438.5342631, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7919288063675048, "elapsed_seconds": 2850.9437029230176, "env_steps_per_second": 10522.831429200647, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3198813439458414, "eta_seconds": 4751.572838205029, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 5.178763572377676e-05, "chaser_grad_norm": 0.14020487293601036, "chaser_new_fields": 80.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.915000915527344, "chaser_subspace_norm": 81.21366119384766, "collisions_per_episode": 2.75, "entropy": 0.0511464960873127, "explorer_grad_norm": 0.7394751831889153, "explorer_new_fields": 78.375, "explorer_partner_vision": 0.0, "explorer_return": 66.73750305175781, "explorer_subspace_norm": 32.37879943847656, "final_distance": 4.293107032775879, "max_param_abs": 0.7756772637367249, "policy_loss": -0.0007455329780867714, "value_loss": 3.648035407066345}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:23:48.563333+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980628.5633435, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.84576081390415, "elapsed_seconds": 3044.7389300549403, "env_steps_per_second": 10509.932291443649, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2686412208562252, "eta_seconds": 4567.10839508241, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0002647830517457095, "chaser_grad_norm": 0.11168603040277958, "chaser_new_fields": 80.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.915000915527344, "chaser_subspace_norm": 72.0607681274414, "collisions_per_episode": 2.3500001430511475, "entropy": 0.051203216426074505, "explorer_grad_norm": 1.5335396528244019, "explorer_new_fields": 79.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 69.2874984741211, "explorer_subspace_norm": 32.942344665527344, "final_distance": 4.0748724937438965, "max_param_abs": 0.8935447335243225, "policy_loss": -0.0006726101947691632, "value_loss": 3.6470033526420593}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:02.358563+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980822.358578, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8986357123963858, "elapsed_seconds": 3235.088564626989, "env_steps_per_second": 10509.758642085353, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2158012579480515, "eta_seconds": 4376.884528612985, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00038216949786851195, "chaser_grad_norm": 0.141121968626976, "chaser_new_fields": 80.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.900001049041748, "chaser_subspace_norm": 79.91463470458984, "collisions_per_episode": 2.325000047683716, "entropy": 0.05178619269281626, "explorer_grad_norm": 1.0006393194198608, "explorer_new_fields": 80.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 70.30000305175781, "explorer_subspace_norm": 35.04092025756836, "final_distance": 3.7615883350372314, "max_param_abs": 0.9136794209480286, "policy_loss": -0.0011855336902684144, "value_loss": 3.6132306456565857}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:12.708195+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981012.708206, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9511827384141442, "elapsed_seconds": 3424.257858290919, "env_steps_per_second": 10513.226950136272, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.162556680283954, "eta_seconds": 4185.204049022234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005947861483002193, "chaser_grad_norm": 0.06370738614350557, "chaser_new_fields": 81.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.255001068115234, "chaser_subspace_norm": 82.45706939697266, "collisions_per_episode": 1.9250000715255737, "entropy": 0.0425002733245492, "explorer_grad_norm": 0.5411928966641426, "explorer_new_fields": 83.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 75.0999984741211, "explorer_subspace_norm": 37.13486099243164, "final_distance": 4.52009916305542, "max_param_abs": 0.9438915848731995, "policy_loss": -0.00047103005892124017, "value_loss": 3.442668378353119}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:21.877490+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981201.877504, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0048930587660936, "elapsed_seconds": 3617.6150115579367, "env_steps_per_second": 10504.158092719543, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.110671275478314, "eta_seconds": 3998.4165917219298, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00011586679148107432, "chaser_grad_norm": 0.037946234457194805, "chaser_new_fields": 81.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.095000743865967, "chaser_subspace_norm": 83.12152862548828, "collisions_per_episode": 2.4000000953674316, "entropy": 0.038095759227871895, "explorer_grad_norm": 0.9206782355904579, "explorer_new_fields": 83.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 74.875, "explorer_subspace_norm": 36.8492431640625, "final_distance": 4.305366516113281, "max_param_abs": 0.9510403871536255, "policy_loss": -0.0005167311039988576, "value_loss": 3.4325848817825317}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:35.234643+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981395.2346568, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0571894302133134, "elapsed_seconds": 3805.8819487679284, "env_steps_per_second": 10510.047483987024, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0571894302133136, "eta_seconds": 3805.881948767929, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -5.382881496068803e-05, "chaser_grad_norm": 0.048281229101121426, "chaser_new_fields": 79.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.890000820159912, "chaser_subspace_norm": 83.30127716064453, "collisions_per_episode": 2.125, "entropy": 0.043126773089170456, "explorer_grad_norm": 1.1048599854111671, "explorer_new_fields": 82.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 73.4124984741211, "explorer_subspace_norm": 39.00022506713867, "final_distance": 4.593392848968506, "max_param_abs": 0.9875046610832214, "policy_loss": -0.0002755683806027198, "value_loss": 3.523101568222046}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:39:43.501580+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981583.5015929, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1102907209330524, "elapsed_seconds": 3997.0465953589883, "env_steps_per_second": 10507.758415617829, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0045487475108568, "eta_seconds": 3616.3754910390844, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006368952735602562, "chaser_grad_norm": 0.22013502568006516, "chaser_new_fields": 79.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.685000896453857, "chaser_subspace_norm": 86.79359436035156, "collisions_per_episode": 2.7750000953674316, "entropy": 0.05470807198435068, "explorer_grad_norm": 0.7701114043593407, "explorer_new_fields": 74.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.375, "explorer_subspace_norm": 37.32197952270508, "final_distance": 5.005620002746582, "max_param_abs": 0.9977173209190369, "policy_loss": -0.0008612548528645902, "value_loss": 3.8843547105789185}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:42:54.666225+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981774.6662369, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.163101664111665, "elapsed_seconds": 4187.165990801994, "env_steps_per_second": 10508.300864273215, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9516286342731803, "eta_seconds": 3425.863083383449, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0003485227200260973, "chaser_grad_norm": 0.04326812829822302, "chaser_new_fields": 82.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.3600006103515625, "chaser_subspace_norm": 93.4374008178711, "collisions_per_episode": 2.2250001430511475, "entropy": 0.034994556568562984, "explorer_grad_norm": 1.0795956999063492, "explorer_new_fields": 74.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 61.900001525878906, "explorer_subspace_norm": 39.19688415527344, "final_distance": 4.504665374755859, "max_param_abs": 1.023972511291504, "policy_loss": -0.000518535818205379, "value_loss": 3.6018770933151245}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:04.785622+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981964.785636, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2151958184602618, "elapsed_seconds": 4374.704946456943, "env_steps_per_second": 10514.994854054152, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8981882136445414, "eta_seconds": 3233.477569120349, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 3.59780934678991e-05, "chaser_grad_norm": 0.11174951866269112, "chaser_new_fields": 82.875, "chaser_partner_vision": 0.0, "chaser_return": 6.490001201629639, "chaser_subspace_norm": 96.23094177246094, "collisions_per_episode": 1.4250000715255737, "entropy": 0.03249161411076784, "explorer_grad_norm": 1.218532234430313, "explorer_new_fields": 81.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 71.76250457763672, "explorer_subspace_norm": 41.40774154663086, "final_distance": 4.823018550872803, "max_param_abs": 1.0638055801391602, "policy_loss": -0.0002506716443460011, "value_loss": 3.340870440006256}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:12.324576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982152.324588, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2680788615552592, "elapsed_seconds": 4565.083901598933, "env_steps_per_second": 10514.593167321167, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8453859077035062, "eta_seconds": 3043.389267732622, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005547765027726914, "chaser_grad_norm": 0.04441197123378515, "chaser_new_fields": 82.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.275001049041748, "chaser_subspace_norm": 91.14032745361328, "collisions_per_episode": 2.799999952316284, "entropy": 0.056606597267091274, "explorer_grad_norm": 0.7461698278784752, "explorer_new_fields": 76.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 64.75, "explorer_subspace_norm": 41.26647186279297, "final_distance": 4.855813026428223, "max_param_abs": 1.091586709022522, "policy_loss": -0.0015027483538716524, "value_loss": 3.5986504554748535}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:22.703534+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982342.7035491, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3209336107683096, "elapsed_seconds": 4755.360998765915, "env_steps_per_second": 10514.448853194473, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7925601664609857, "eta_seconds": 2853.2165992595487, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00023565270424291684, "chaser_grad_norm": 0.0892906654626131, "chaser_new_fields": 82.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.360001564025879, "chaser_subspace_norm": 91.74296569824219, "collisions_per_episode": 1.9250000715255737, "entropy": 0.03169793915003538, "explorer_grad_norm": 0.5501591935753822, "explorer_new_fields": 78.375, "explorer_partner_vision": 0.0, "explorer_return": 66.92500305175781, "explorer_subspace_norm": 41.62643814086914, "final_distance": 4.7033281326293945, "max_param_abs": 1.072445273399353, "policy_loss": -0.00034868576267022, "value_loss": 3.528638005256653}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:32.980629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982532.9806402, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3748724833491723, "elapsed_seconds": 4949.540940057021, "env_steps_per_second": 10506.02482730468, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7403159525726312, "eta_seconds": 2665.1374292614723, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0004677979677508437, "chaser_grad_norm": 0.09906978532671928, "chaser_new_fields": 82.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.345000743865967, "chaser_subspace_norm": 83.64076232910156, "collisions_per_episode": 1.75, "entropy": 0.018137681763619184, "explorer_grad_norm": 0.82523462921381, "explorer_new_fields": 79.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 69.2125015258789, "explorer_subspace_norm": 41.203739166259766, "final_distance": 4.766340255737305, "max_param_abs": 1.1244043111801147, "policy_loss": -0.000598826434694999, "value_loss": 3.4963014125823975}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:58:47.160572+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982727.160585, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4286773643683086, "elapsed_seconds": 5143.238511725911, "env_steps_per_second": 10499.221429627863, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6878816939551116, "eta_seconds": 2476.3740982384015, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -7.345536402177111e-05, "chaser_grad_norm": 0.06949483370408416, "chaser_new_fields": 83.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.565001010894775, "chaser_subspace_norm": 95.91769409179688, "collisions_per_episode": 1.975000023841858, "entropy": 0.034317951649427414, "explorer_grad_norm": 0.7347743660211563, "explorer_new_fields": 78.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.70000457763672, "explorer_subspace_norm": 41.50311279296875, "final_distance": 5.217411994934082, "max_param_abs": 1.160978078842163, "policy_loss": -0.0004958413387612381, "value_loss": 3.5299025774002075}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:00.858142+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982920.8581529, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.482456398446391, "elapsed_seconds": 5336.8430344070075, "env_steps_per_second": 10493.094820095703, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6353384564770247, "eta_seconds": 2287.218443317289, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -2.9276243399545265e-06, "chaser_grad_norm": 0.07764142286032438, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.4850006103515625, "chaser_subspace_norm": 100.48745727539062, "collisions_per_episode": 2.575000047683716, "entropy": 0.017260680440813303, "explorer_grad_norm": 1.8790609538555145, "explorer_new_fields": 81.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 71.8375015258789, "explorer_subspace_norm": 42.31975555419922, "final_distance": 5.080520153045654, "max_param_abs": 1.1950066089630127, "policy_loss": -0.00023038364184202464, "value_loss": 3.3809183835983276}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:14.462666+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983114.4626808, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5346350495336163, "elapsed_seconds": 5524.686178321019, "env_steps_per_second": 10498.333865114942, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5821029498230958, "eta_seconds": 2095.570619363145, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -2.511701886417049e-05, "chaser_grad_norm": 0.0464151119813323, "chaser_new_fields": 83.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.67000150680542, "chaser_subspace_norm": 95.02689361572266, "collisions_per_episode": 1.899999976158142, "entropy": 0.00915247411467135, "explorer_grad_norm": 0.5128177627921104, "explorer_new_fields": 80.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 70.07500457763672, "explorer_subspace_norm": 43.29038619995117, "final_distance": 4.938743591308594, "max_param_abs": 1.2273591756820679, "policy_loss": -5.457508942985356e-05, "value_loss": 3.422726094722748}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:22.305809+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983302.305819, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5866568446219511, "elapsed_seconds": 5711.964640639024, "env_steps_per_second": 10504.266705910057, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5288856148739837, "eta_seconds": 1903.9882135463413, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": -2.877289655384835e-05, "chaser_grad_norm": 0.24285456165671349, "chaser_new_fields": 83.125, "chaser_partner_vision": 0.0, "chaser_return": 6.525001049041748, "chaser_subspace_norm": 92.94022369384766, "collisions_per_episode": 2.075000047683716, "entropy": 0.029723917599767447, "explorer_grad_norm": 0.9983538910746574, "explorer_new_fields": 79.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.0250015258789, "explorer_subspace_norm": 42.98257827758789, "final_distance": 4.934007167816162, "max_param_abs": 1.2273591756820679, "policy_loss": -0.0005432382828587379, "value_loss": 3.5983139276504517}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:29.584271+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983489.584286, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6405488199960867, "elapsed_seconds": 5905.975751985912, "env_steps_per_second": 10497.841949173633, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4762883670956381, "eta_seconds": 1714.638121544297, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0003411350088956093, "chaser_grad_norm": 0.14202416315674782, "chaser_new_fields": 81.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.2400007247924805, "chaser_subspace_norm": 103.27901458740234, "collisions_per_episode": 2.200000047683716, "entropy": 0.024732754565775394, "explorer_grad_norm": 1.4782511293888092, "explorer_new_fields": 80.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 70.0, "explorer_subspace_norm": 40.86754608154297, "final_distance": 4.7141900062561035, "max_param_abs": 1.280949354171753, "policy_loss": -0.0005018190426042679, "value_loss": 3.4643369913101196}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:14:43.595384+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983683.5953975, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6925805555147235, "elapsed_seconds": 6093.289999853005, "env_steps_per_second": 10503.356971610401, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4231451388786809, "eta_seconds": 1523.3224999632512, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 3.3645105060708946e-05, "chaser_grad_norm": 0.15357054583728313, "chaser_new_fields": 83.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.6100006103515625, "chaser_subspace_norm": 99.26699829101562, "collisions_per_episode": 1.5, "entropy": 0.031850278843194246, "explorer_grad_norm": 1.5429618656635284, "explorer_new_fields": 80.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 70.9375, "explorer_subspace_norm": 38.98711013793945, "final_distance": 5.134640693664551, "max_param_abs": 1.2991660833358765, "policy_loss": -0.00012170585207105233, "value_loss": 3.484034240245819}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:17:50.909631+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983870.9096456, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.746258258684991, "elapsed_seconds": 6286.529731265968, "env_steps_per_second": 10498.638011962295, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3704184185089375, "eta_seconds": 1333.506306632175, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": -0.00012888822604309458, "chaser_grad_norm": 0.07348597142845392, "chaser_new_fields": 82.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.345000743865967, "chaser_subspace_norm": 101.37876892089844, "collisions_per_episode": 2.0, "entropy": 0.035380104556679726, "explorer_grad_norm": 0.5114077851176262, "explorer_new_fields": 80.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 70.1500015258789, "explorer_subspace_norm": 40.77675247192383, "final_distance": 4.24074649810791, "max_param_abs": 1.3488653898239136, "policy_loss": -0.0001923652420892097, "value_loss": 3.485174000263214}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:04.149364+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984064.1493769, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8001988725216425, "elapsed_seconds": 6480.715941077913, "env_steps_per_second": 10492.667880871479, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3176821539744075, "eta_seconds": 1143.655754307867, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0005805552811085635, "chaser_grad_norm": 0.07853336445987225, "chaser_new_fields": 81.5, "chaser_partner_vision": 0.0, "chaser_return": 6.195001125335693, "chaser_subspace_norm": 97.96226501464844, "collisions_per_episode": 2.1000001430511475, "entropy": 0.055465965531766415, "explorer_grad_norm": 0.5799347162246704, "explorer_new_fields": 79.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.2125015258789, "explorer_subspace_norm": 41.3720588684082, "final_distance": 4.8930487632751465, "max_param_abs": 1.3339811563491821, "policy_loss": -0.001213232978015455, "value_loss": 3.599591374397278}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:18.335574+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984258.3355885, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8534982745891708, "elapsed_seconds": 6672.593788521015, "env_steps_per_second": 10490.673075352239, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26478546779845297, "eta_seconds": 953.2276840744307, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00023692901135386434, "chaser_grad_norm": 0.09657379426062107, "chaser_new_fields": 83.125, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 95.30155944824219, "collisions_per_episode": 1.9250000715255737, "entropy": 0.043237156234681606, "explorer_grad_norm": 0.8380573987960815, "explorer_new_fields": 78.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 66.92500305175781, "explorer_subspace_norm": 40.663387298583984, "final_distance": 5.157171726226807, "max_param_abs": 1.3518613576889038, "policy_loss": -0.0006293986412941877, "value_loss": 3.5378934144973755}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:30.213419+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984450.2134297, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9066078050941644, "elapsed_seconds": 6863.788098338991, "env_steps_per_second": 10489.834325949501, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21184531167712933, "eta_seconds": 762.6431220376656, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0008987359628854635, "chaser_grad_norm": 0.22547512501478195, "chaser_new_fields": 80.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.885000705718994, "chaser_subspace_norm": 93.97661590576172, "collisions_per_episode": 1.975000023841858, "entropy": 0.05276448093354702, "explorer_grad_norm": 0.39018620550632477, "explorer_new_fields": 79.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 69.13750457763672, "explorer_subspace_norm": 39.80522537231445, "final_distance": 5.034396648406982, "max_param_abs": 1.3761506080627441, "policy_loss": -0.0011541230918368228, "value_loss": 3.5250176191329956}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:30:41.407731+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984641.407746, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9588988603477548, "elapsed_seconds": 7052.035897251917, "env_steps_per_second": 10493.423612440316, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15882963732549366, "eta_seconds": 571.7866943717771, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00029414826620524637, "chaser_grad_norm": 0.25725705921649933, "chaser_new_fields": 80.25, "chaser_partner_vision": 0.0, "chaser_return": 5.940001010894775, "chaser_subspace_norm": 97.96788787841797, "collisions_per_episode": 4.849999904632568, "entropy": 0.040001179091632366, "explorer_grad_norm": 0.5714121311903, "explorer_new_fields": 78.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 66.55000305175781, "explorer_subspace_norm": 41.226619720458984, "final_distance": 4.181025981903076, "max_param_abs": 1.4031187295913696, "policy_loss": -0.001016252324102851, "value_loss": 3.6369670629501343}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:33:49.655527+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984829.6555393, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.010999525518047, "elapsed_seconds": 7239.598291864968, "env_steps_per_second": 10497.820035871342, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1058420802904235, "eta_seconds": 381.0314890455246, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0003797909571186253, "chaser_grad_norm": 0.18442872539162636, "chaser_new_fields": 82.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.305001258850098, "chaser_subspace_norm": 96.6340103149414, "collisions_per_episode": 2.4000000953674316, "entropy": 0.046740551479160786, "explorer_grad_norm": 1.5118907243013382, "explorer_new_fields": 80.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.2874984741211, "explorer_subspace_norm": 40.9044189453125, "final_distance": 4.769104480743408, "max_param_abs": 1.5274485349655151, "policy_loss": -0.00024989158625032815, "value_loss": 3.4859007000923157}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:36:57.217923+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985017.2179375, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.0636419246980546, "elapsed_seconds": 7429.110928912996, "env_steps_per_second": 10499.237492394357, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05291389550507832, "eta_seconds": 190.49002381828197, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00011672284622749052, "chaser_grad_norm": 0.03770865499973297, "chaser_new_fields": 83.0, "chaser_partner_vision": 0.0, "chaser_return": 6.4850006103515625, "chaser_subspace_norm": 90.64308166503906, "collisions_per_episode": 2.424999952316284, "entropy": 0.02636505337432027, "explorer_grad_norm": 2.2632810473442078, "explorer_new_fields": 82.125, "explorer_partner_vision": 0.0, "explorer_return": 72.51250457763672, "explorer_subspace_norm": 40.891624450683594, "final_distance": 5.000566005706787, "max_param_abs": 1.5338480472564697, "policy_loss": -0.000305423214598477, "value_loss": 3.3947563767433167}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:06.730559+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985206.730569, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1173314844191515, "elapsed_seconds": 7622.393343908945, "env_steps_per_second": 10495.3911967726, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:20.012977+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985400.0129912, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1173361086988653, "elapsed_seconds": 7622.409991315915, "env_steps_per_second": 10495.368274750725, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:20.029624+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985400.0296402, "update": 20000} diff --git a/non_social/seed_0002/attempt_01/status.json b/non_social/seed_0002/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..1a2490edf4e6cda20f5e3139da7a7eb42e1ef088 --- /dev/null +++ b/non_social/seed_0002/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1173361086988653, + "elapsed_seconds": 7622.409991315915, + "env_steps_per_second": 10495.368274750725, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.00015983218477721994, + "chaser_grad_norm": 0.06889200676232576, + "chaser_new_fields": 84.3499984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.7600016593933105, + "chaser_subspace_norm": 98.9588623046875, + "collisions_per_episode": 2.075000047683716, + "entropy": 0.02094564540311694, + "explorer_grad_norm": 1.1887003779411316, + "explorer_new_fields": 82.75, + "explorer_partner_vision": 0.0, + "explorer_return": 73.30000305175781, + "explorer_subspace_norm": 38.70703887939453, + "final_distance": 4.667576789855957, + "max_param_abs": 1.5199339389801025, + "policy_loss": -0.0006238509232616707, + "value_loss": 3.509905695915222 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:20.029624+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985400.0296402, + "update": 20000 +} diff --git a/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ddd6ac101c470bae435c012541682211b094b6b4 --- /dev/null +++ b/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:17f68c1c23df5485beda4cdb0392ee32910c36897733023f84588c380da48191 +size 11460384 diff --git a/non_social/seed_0003/attempt_01/config.json b/non_social/seed_0003/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..b0b5dadefa457632d0c1eee6c5960ccfdd9841aa --- /dev/null +++ b/non_social/seed_0003/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", + "seed": 3, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0003/attempt_01/metrics.jsonl b/non_social/seed_0003/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..06a75f6524e4afec7d04e02a205f0933f22e17c8 --- /dev/null +++ b/non_social/seed_0003/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005703172285898795, "elapsed_seconds": 2.0531420229235664, "env_steps_per_second": 1948.2334662383512, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 11.405774254569002, "eta_seconds": 41060.787316448404, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 3.7429647265534854e-05, "chaser_grad_norm": 0.4683794528245926, "chaser_new_fields": 31.42500114440918, "chaser_partner_vision": 0.0, "chaser_return": -3.7850005626678467, "chaser_subspace_norm": 0.002194158500060439, "collisions_per_episode": 2.1000001430511475, "entropy": 2.7725386023521423, "explorer_grad_norm": 0.12834803014993668, "explorer_new_fields": 33.17499923706055, "explorer_partner_vision": 0.0, "explorer_return": -0.574999988079071, "explorer_subspace_norm": 0.00197780504822731, "final_distance": 4.789273262023926, "max_param_abs": 0.1542910486459732, "policy_loss": -0.0004491267254955744, "value_loss": 3.2673333883285522}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.784750+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.7847638, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05423463147389056, "elapsed_seconds": 195.244673306006, "env_steps_per_second": 10243.557307529767, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.115150627481732, "eta_seconds": 7614.542258934235, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0008568063229401623, "chaser_grad_norm": 0.18940076977014542, "chaser_new_fields": 76.25, "chaser_partner_vision": 0.0, "chaser_return": 5.140000820159912, "chaser_subspace_norm": 15.807612419128418, "collisions_per_episode": 3.299999952316284, "entropy": 0.5341657549142838, "explorer_grad_norm": 0.46438099443912506, "explorer_new_fields": 66.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 48.54999923706055, "explorer_subspace_norm": 7.001035213470459, "final_distance": 6.011678695678711, "max_param_abs": 0.3208908438682556, "policy_loss": -0.0024413161915717296, "value_loss": 4.51411771774292}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:32.976281+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977972.9762921, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10752084387055624, "elapsed_seconds": 387.07503793400247, "env_steps_per_second": 10333.913603289533, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0428960335405684, "eta_seconds": 7354.425720746047, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0003562065489146615, "chaser_grad_norm": 0.18975985795259476, "chaser_new_fields": 78.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.505001068115234, "chaser_subspace_norm": 23.10985565185547, "collisions_per_episode": 4.099999904632568, "entropy": 0.40523532032966614, "explorer_grad_norm": 0.9458442181348801, "explorer_new_fields": 65.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 46.67499923706055, "explorer_subspace_norm": 7.960142612457275, "final_distance": 4.668750286102295, "max_param_abs": 0.3501768112182617, "policy_loss": -0.0017710868579001726, "value_loss": 4.498179197311401}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:44.806645+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978164.8066587, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15951738449887165, "elapsed_seconds": 574.262584195938, "env_steps_per_second": 10448.182007889276, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9673810754860839, "eta_seconds": 7082.571871749902, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 6.207642075262143e-05, "chaser_grad_norm": 0.40830305963754654, "chaser_new_fields": 80.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.0450005531311035, "chaser_subspace_norm": 32.19874572753906, "collisions_per_episode": 2.4750001430511475, "entropy": 0.27992892265319824, "explorer_grad_norm": 0.4816494658589363, "explorer_new_fields": 62.67499923706055, "explorer_partner_vision": 0.0, "explorer_return": 43.337501525878906, "explorer_subspace_norm": 9.728968620300293, "final_distance": 5.00688362121582, "max_param_abs": 0.37524423003196716, "policy_loss": -0.0013687876923853537, "value_loss": 4.304821252822876}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:51.994189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978351.9942012, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21146941066053437, "elapsed_seconds": 761.2898783779237, "env_steps_per_second": 10508.480707829134, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9032246959448094, "eta_seconds": 6851.608905401314, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0005010455043041873, "chaser_grad_norm": 0.15113239735364914, "chaser_new_fields": 80.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.935000896453857, "chaser_subspace_norm": 39.81361389160156, "collisions_per_episode": 2.325000047683716, "entropy": 0.18511592969298363, "explorer_grad_norm": 1.2664929926395416, "explorer_new_fields": 61.92499923706055, "explorer_partner_vision": 0.0, "explorer_return": 42.212501525878906, "explorer_subspace_norm": 12.132214546203613, "final_distance": 5.070174217224121, "max_param_abs": 0.41063612699508667, "policy_loss": -0.0010882803588145862, "value_loss": 4.197474360466003}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:48:59.021486+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978539.0214996, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2641038031524901, "elapsed_seconds": 950.7736913489643, "env_steps_per_second": 10517.750008219022, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8487266220674308, "eta_seconds": 6655.41583944275, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 6.692116872819565e-05, "chaser_grad_norm": 0.10173061676323414, "chaser_new_fields": 82.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.280001163482666, "chaser_subspace_norm": 44.502197265625, "collisions_per_episode": 2.375, "entropy": 0.2088600993156433, "explorer_grad_norm": 0.4088284820318222, "explorer_new_fields": 62.60000228881836, "explorer_partner_vision": 0.0, "explorer_return": 43.22500228881836, "explorer_subspace_norm": 11.41265869140625, "final_distance": 5.094779968261719, "max_param_abs": 0.4990295469760895, "policy_loss": -0.00045264054241300755, "value_loss": 4.0946797132492065}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:08.505299+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978728.5053127, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3166092689913866, "elapsed_seconds": 1139.7933683689917, "env_steps_per_second": 10528.224091329485, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7941191909511907, "eta_seconds": 6458.829087424287, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00024391082534200303, "chaser_grad_norm": 0.12368607148528099, "chaser_new_fields": 82.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.370000839233398, "chaser_subspace_norm": 46.88285446166992, "collisions_per_episode": 2.325000047683716, "entropy": 0.1747008040547371, "explorer_grad_norm": 0.6444325372576714, "explorer_new_fields": 62.525001525878906, "explorer_partner_vision": 0.0, "explorer_return": 43.07500076293945, "explorer_subspace_norm": 12.49425220489502, "final_distance": 5.174897193908691, "max_param_abs": 0.5542518496513367, "policy_loss": -0.0005140939082473039, "value_loss": 3.8949413895606995}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:17.524975+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978917.5249884, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3700827708974895, "elapsed_seconds": 1332.297975230962, "env_steps_per_second": 10508.159781277918, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7446759199453075, "eta_seconds": 6280.833311803107, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00018153257259567734, "chaser_grad_norm": 0.03313362039625645, "chaser_new_fields": 83.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.580000400543213, "chaser_subspace_norm": 54.41621017456055, "collisions_per_episode": 2.5, "entropy": 0.1896023191511631, "explorer_grad_norm": 1.3531986810266972, "explorer_new_fields": 64.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 46.07500076293945, "explorer_subspace_norm": 12.994067192077637, "final_distance": 6.120728492736816, "max_param_abs": 0.6412327885627747, "policy_loss": -0.0006707920666073619, "value_loss": 3.8537389636039734}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:30.029581+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979110.029594, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.423166916928862, "elapsed_seconds": 1523.4009009439033, "env_steps_per_second": 10502.816422181684, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.692667667715448, "eta_seconds": 6093.603603775613, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": -0.00017728790872564848, "chaser_grad_norm": 0.06281640473753214, "chaser_new_fields": 83.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.6100006103515625, "chaser_subspace_norm": 66.46906280517578, "collisions_per_episode": 2.2750000953674316, "entropy": 0.13005392998456955, "explorer_grad_norm": 0.7341135591268539, "explorer_new_fields": 64.0, "explorer_partner_vision": 0.0, "explorer_return": 45.47500228881836, "explorer_subspace_norm": 14.190570831298828, "final_distance": 5.387036323547363, "max_param_abs": 0.7567383646965027, "policy_loss": -0.00035322338224119676, "value_loss": 3.765820026397705}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:41.132508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979301.1325223, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4764483611155366, "elapsed_seconds": 1715.2141000159318, "env_steps_per_second": 10494.316715232697, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6410999105090707, "eta_seconds": 5907.959677832654, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00019356987170110473, "chaser_grad_norm": 0.07370109297335148, "chaser_new_fields": 82.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.375000953674316, "chaser_subspace_norm": 64.52543640136719, "collisions_per_episode": 1.9250000715255737, "entropy": 0.09183669462800026, "explorer_grad_norm": 0.9774608314037323, "explorer_new_fields": 63.45000076293945, "explorer_partner_vision": 0.0, "explorer_return": 44.72500228881836, "explorer_subspace_norm": 16.661361694335938, "final_distance": 5.7826080322265625, "max_param_abs": 0.8368179798126221, "policy_loss": -0.0005020306385699058, "value_loss": 3.812152862548828}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:04:52.945707+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979492.9457178, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.530420549550553, "elapsed_seconds": 1909.5139783819905, "env_steps_per_second": 10473.869385835458, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5912616486516586, "eta_seconds": 5728.541935145971, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00014348731071001097, "chaser_grad_norm": 0.05367972236126661, "chaser_new_fields": 84.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.710001468658447, "chaser_subspace_norm": 71.2568359375, "collisions_per_episode": 1.725000023841858, "entropy": 0.09447645954787731, "explorer_grad_norm": 1.0041202381253242, "explorer_new_fields": 63.04999923706055, "explorer_partner_vision": 0.0, "explorer_return": 43.97500228881836, "explorer_subspace_norm": 17.77263641357422, "final_distance": 6.084954261779785, "max_param_abs": 0.9192169308662415, "policy_loss": -0.0004371514491481321, "value_loss": 3.8562639355659485}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:07.245587+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979687.245602, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5843529775638793, "elapsed_seconds": 2103.6707192299655, "env_steps_per_second": 10457.910450953537, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5405669408502274, "eta_seconds": 5546.040987060818, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 3.95765696028505e-05, "chaser_grad_norm": 0.07043039426207542, "chaser_new_fields": 82.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.345000743865967, "chaser_subspace_norm": 69.50054168701172, "collisions_per_episode": 2.3500001430511475, "entropy": 0.06500965170562267, "explorer_grad_norm": 0.3183881863951683, "explorer_new_fields": 65.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 47.01250076293945, "explorer_subspace_norm": 18.546667098999023, "final_distance": 6.202277660369873, "max_param_abs": 0.9679896831512451, "policy_loss": -0.0005637256332065022, "value_loss": 3.658661365509033}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:21.402325+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979881.4023366, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.638017752329696, "elapsed_seconds": 2296.8639083869057, "env_steps_per_second": 10449.030050219768, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4887080887692905, "eta_seconds": 5359.349119569446, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00022892639110892732, "chaser_grad_norm": 0.07459682039916515, "chaser_new_fields": 83.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.495001316070557, "chaser_subspace_norm": 76.20964813232422, "collisions_per_episode": 2.25, "entropy": 0.07609122060239315, "explorer_grad_norm": 1.088119976222515, "explorer_new_fields": 66.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 49.375, "explorer_subspace_norm": 18.288787841796875, "final_distance": 5.6315436363220215, "max_param_abs": 0.9871170520782471, "policy_loss": -0.0005530785417064976, "value_loss": 3.732019543647766}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:34.595515+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980074.5955298, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6912928554783058, "elapsed_seconds": 2488.654279721901, "env_steps_per_second": 10447.413371898896, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4357620844549426, "eta_seconds": 5168.743504037793, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00019669308521230722, "chaser_grad_norm": 0.176544526591897, "chaser_new_fields": 81.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.260000705718994, "chaser_subspace_norm": 76.09440612792969, "collisions_per_episode": 1.8250000476837158, "entropy": 0.0995361227542162, "explorer_grad_norm": 1.0598936080932617, "explorer_new_fields": 64.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 45.66250228881836, "explorer_subspace_norm": 19.796846389770508, "final_distance": 6.028972148895264, "max_param_abs": 1.0148918628692627, "policy_loss": -0.0004410063167483713, "value_loss": 3.905380606651306}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:46.385887+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980266.3859003, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7447569418966627, "elapsed_seconds": 2681.124990827986, "env_steps_per_second": 10443.377349354023, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3831200349509452, "eta_seconds": 4979.232125823402, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0006360402063246562, "chaser_grad_norm": 0.07907636929303408, "chaser_new_fields": 83.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.520001411437988, "chaser_subspace_norm": 74.84884643554688, "collisions_per_episode": 2.2750000953674316, "entropy": 0.06973402015864849, "explorer_grad_norm": 0.9930750504136086, "explorer_new_fields": 67.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 49.9375, "explorer_subspace_norm": 22.50824737548828, "final_distance": 5.44460391998291, "max_param_abs": 1.0148918628692627, "policy_loss": -0.0013815705683093427, "value_loss": 3.6661624908447266}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:20:58.856599+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980458.8566139, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7980109253722023, "elapsed_seconds": 2872.8393313399283, "env_steps_per_second": 10442.630631211674, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3300182089536705, "eta_seconds": 4788.065552233214, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 5.969263143233705e-05, "chaser_grad_norm": 0.04893398657441139, "chaser_new_fields": 85.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.915000915527344, "chaser_subspace_norm": 82.32737731933594, "collisions_per_episode": 1.725000023841858, "entropy": 0.04560158494859934, "explorer_grad_norm": 0.8938560858368874, "explorer_new_fields": 72.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 58.75, "explorer_subspace_norm": 24.45636749267578, "final_distance": 5.509303569793701, "max_param_abs": 1.0297765731811523, "policy_loss": -0.0006032357893133167, "value_loss": 3.484142780303955}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:10.570937+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980650.5709474, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8505744172652744, "elapsed_seconds": 3062.0679021549877, "env_steps_per_second": 10450.454079571326, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2758616258979116, "eta_seconds": 4593.1018532324815, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0002857226364572085, "chaser_grad_norm": 0.14095772989094257, "chaser_new_fields": 83.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.6100006103515625, "chaser_subspace_norm": 85.53858947753906, "collisions_per_episode": 2.674999952316284, "entropy": 0.027820547111332417, "explorer_grad_norm": 0.5691776648163795, "explorer_new_fields": 71.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 56.5, "explorer_subspace_norm": 25.945280075073242, "final_distance": 5.516311168670654, "max_param_abs": 1.023421287536621, "policy_loss": -0.0003698255095514469, "value_loss": 3.6125537753105164}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:19.799508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980839.7995217, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9034056384319168, "elapsed_seconds": 3252.2602983549004, "env_steps_per_second": 10454.267764852128, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2222546872902404, "eta_seconds": 4400.116874244865, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 5.766702054116335e-05, "chaser_grad_norm": 0.25868288427591324, "chaser_new_fields": 84.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.835000514984131, "chaser_subspace_norm": 89.00960540771484, "collisions_per_episode": 1.5, "entropy": 0.04216903727501631, "explorer_grad_norm": 0.4825250655412674, "explorer_new_fields": 70.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 55.45000076293945, "explorer_subspace_norm": 25.861480712890625, "final_distance": 5.273017883300781, "max_param_abs": 1.0302062034606934, "policy_loss": -0.0005198416978819953, "value_loss": 3.5706403255462646}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:29.991904+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981029.991915, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9566017068602701, "elapsed_seconds": 3443.7661446969723, "env_steps_per_second": 10453.671500149947, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1691798639403304, "eta_seconds": 4209.047510185189, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00012291349896573212, "chaser_grad_norm": 0.1024897675961256, "chaser_new_fields": 85.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.970001220703125, "chaser_subspace_norm": 90.3985595703125, "collisions_per_episode": 1.600000023841858, "entropy": 0.056081829592585564, "explorer_grad_norm": 0.9510652124881744, "explorer_new_fields": 71.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 57.17499923706055, "explorer_subspace_norm": 26.2315673828125, "final_distance": 6.142429828643799, "max_param_abs": 1.0548176765441895, "policy_loss": -0.0005760992522567676, "value_loss": 3.5523917078971863}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:41.497751+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981221.4977636, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.009516894095836, "elapsed_seconds": 3634.2608187450096, "env_steps_per_second": 10456.046468652252, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1157818303164504, "eta_seconds": 4016.8145891392214, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0002783891020918805, "chaser_grad_norm": 0.04646790958940983, "chaser_new_fields": 85.625, "chaser_partner_vision": 0.0, "chaser_return": 7.0100016593933105, "chaser_subspace_norm": 86.5994873046875, "collisions_per_episode": 1.625, "entropy": 0.05779918283224106, "explorer_grad_norm": 0.4065385088324547, "explorer_new_fields": 72.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 58.22500228881836, "explorer_subspace_norm": 26.266204833984375, "final_distance": 5.877896785736084, "max_param_abs": 1.0548176765441895, "policy_loss": -0.000716598862897655, "value_loss": 3.518312096595764}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:51.992426+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981411.9924395, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0632194601316587, "elapsed_seconds": 3827.5900564739713, "env_steps_per_second": 10450.43993996801, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0632194601316587, "eta_seconds": 3827.5900564739713, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0003546991464112581, "chaser_grad_norm": 0.03811489138752222, "chaser_new_fields": 84.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.770001411437988, "chaser_subspace_norm": 84.1824951171875, "collisions_per_episode": 1.649999976158142, "entropy": 0.05527436640113592, "explorer_grad_norm": 1.8407804071903229, "explorer_new_fields": 69.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 54.13750076293945, "explorer_subspace_norm": 25.939453125, "final_distance": 5.767298221588135, "max_param_abs": 1.0759329795837402, "policy_loss": -0.0006026283805837807, "value_loss": 3.7109662294387817}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:40:05.321665+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981605.32168, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.115201353638614, "elapsed_seconds": 4014.7248730990104, "env_steps_per_second": 10461.488975601891, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.008991700911127, "eta_seconds": 3632.370123280057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0002795625753169162, "chaser_grad_norm": 0.08300566114485264, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.750000953674316, "chaser_subspace_norm": 90.17914581298828, "collisions_per_episode": 2.0, "entropy": 0.048600868321955204, "explorer_grad_norm": 0.7562909424304962, "explorer_new_fields": 70.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 54.8125, "explorer_subspace_norm": 25.51056671142578, "final_distance": 5.623228073120117, "max_param_abs": 1.0979514122009277, "policy_loss": -0.0011265874402646148, "value_loss": 3.518922805786133}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:12.456478+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981792.4564893, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1669910882661094, "elapsed_seconds": 4201.167917757994, "env_steps_per_second": 10473.278112501906, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9548108903995439, "eta_seconds": 3437.319205438358, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": -2.8862833428219825e-05, "chaser_grad_norm": 0.11567963287234306, "chaser_new_fields": 84.75, "chaser_partner_vision": 0.0, "chaser_return": 6.850000858306885, "chaser_subspace_norm": 97.20673370361328, "collisions_per_episode": 1.5, "entropy": 0.0665871873497963, "explorer_grad_norm": 1.5792032703757286, "explorer_new_fields": 71.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.17499923706055, "explorer_subspace_norm": 25.10002326965332, "final_distance": 5.25277042388916, "max_param_abs": 1.1487668752670288, "policy_loss": 4.225992062245609e-05, "value_loss": 3.6418476700782776}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:18.899524+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981978.8995383, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2197464188535942, "elapsed_seconds": 4391.08710787294, "env_steps_per_second": 10475.76576595917, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.901551700891787, "eta_seconds": 3245.5861232104335, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0001558729046919477, "chaser_grad_norm": 0.11555503867566586, "chaser_new_fields": 85.625, "chaser_partner_vision": 0.0, "chaser_return": 7.04500150680542, "chaser_subspace_norm": 87.86775207519531, "collisions_per_episode": 1.375, "entropy": 0.07149608805775642, "explorer_grad_norm": 1.9915097206830978, "explorer_new_fields": 70.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 54.32500076293945, "explorer_subspace_norm": 24.941837310791016, "final_distance": 5.770956516265869, "max_param_abs": 1.2198792695999146, "policy_loss": -0.0006176797202748219, "value_loss": 3.390411138534546}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:28.818715+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982168.8187284, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2738369098308289, "elapsed_seconds": 4585.812875390984, "env_steps_per_second": 10467.064684994924, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8492246065538859, "eta_seconds": 3057.2085835939893, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00026339444551005953, "chaser_grad_norm": 0.16070806607604027, "chaser_new_fields": 86.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.185001373291016, "chaser_subspace_norm": 83.01050567626953, "collisions_per_episode": 1.774999976158142, "entropy": 0.034742836840450764, "explorer_grad_norm": 0.5259442552924156, "explorer_new_fields": 69.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 54.13750076293945, "explorer_subspace_norm": 27.409799575805664, "final_distance": 6.103976726531982, "max_param_abs": 1.2449547052383423, "policy_loss": -0.0005283119698304972, "value_loss": 3.4610748887062073}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:43.544482+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982363.5444918, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3272883638097037, "elapsed_seconds": 4778.238109714934, "env_steps_per_second": 10464.108077481924, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7963730182858223, "eta_seconds": 2866.9428658289603, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00010838919917051382, "chaser_grad_norm": 0.051509371027350426, "chaser_new_fields": 85.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.080001354217529, "chaser_subspace_norm": 87.01652526855469, "collisions_per_episode": 1.5, "entropy": 0.04255330376327038, "explorer_grad_norm": 0.8837704248726368, "explorer_new_fields": 69.375, "explorer_partner_vision": 0.0, "explorer_return": 53.38750076293945, "explorer_subspace_norm": 27.55086326599121, "final_distance": 5.9578986167907715, "max_param_abs": 1.2449547052383423, "policy_loss": -0.0006131164104443698, "value_loss": 3.4126020073890686}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:55.969715+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982555.9697256, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3816036467694277, "elapsed_seconds": 4973.7731283699395, "env_steps_per_second": 10454.839546942107, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.743940425183538, "eta_seconds": 2678.1855306607367, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00014906735077543226, "chaser_grad_norm": 0.02636421099305153, "chaser_new_fields": 86.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 7.185000896453857, "chaser_subspace_norm": 85.54673767089844, "collisions_per_episode": 1.6750000715255737, "entropy": 0.05380332376807928, "explorer_grad_norm": 0.4192870259284973, "explorer_new_fields": 69.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 53.20000076293945, "explorer_subspace_norm": 26.98349380493164, "final_distance": 6.223194599151611, "max_param_abs": 1.264591932296753, "policy_loss": -0.0005827080261688877, "value_loss": 3.5476850271224976}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:11.504737+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982751.5047536, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4348082006950023, "elapsed_seconds": 5165.309522502008, "env_steps_per_second": 10454.358981733027, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6908335781124085, "eta_seconds": 2487.000881204671, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 4.812631965930153e-05, "chaser_grad_norm": 0.02265438949689269, "chaser_new_fields": 86.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.200000762939453, "chaser_subspace_norm": 91.70532989501953, "collisions_per_episode": 2.0, "entropy": 0.05330873932689428, "explorer_grad_norm": 0.6988201290369034, "explorer_new_fields": 67.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 50.837501525878906, "explorer_subspace_norm": 28.826154708862305, "final_distance": 5.746100902557373, "max_param_abs": 1.3166861534118652, "policy_loss": -0.00022317440832608781, "value_loss": 3.510417342185974}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:23.041130+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982943.041143, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4877512854633257, "elapsed_seconds": 5355.904627667973, "env_steps_per_second": 10455.750035336812, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6376076937699967, "eta_seconds": 2295.387697571988, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 3.361281381987491e-05, "chaser_grad_norm": 0.16800860688090324, "chaser_new_fields": 85.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.975000858306885, "chaser_subspace_norm": 90.61492156982422, "collisions_per_episode": 3.5, "entropy": 0.04608313553035259, "explorer_grad_norm": 0.5840689986944199, "explorer_new_fields": 67.375, "explorer_partner_vision": 0.0, "explorer_return": 50.20000076293945, "explorer_subspace_norm": 28.46355438232422, "final_distance": 4.918022155761719, "max_param_abs": 1.3287265300750732, "policy_loss": -0.000655316377524251, "value_loss": 3.713947057723999}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:33.636235+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983133.6362493, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5396821022677534, "elapsed_seconds": 5542.855568163912, "env_steps_per_second": 10463.920498511687, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5840173491360444, "eta_seconds": 2102.46245688976, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0002454529702964059, "chaser_grad_norm": 0.038383228704333305, "chaser_new_fields": 81.75, "chaser_partner_vision": 0.0, "chaser_return": 6.220000743865967, "chaser_subspace_norm": 85.26435089111328, "collisions_per_episode": 2.125, "entropy": 0.051656256429851055, "explorer_grad_norm": 1.2297004908323288, "explorer_new_fields": 64.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 46.07500076293945, "explorer_subspace_norm": 28.52336883544922, "final_distance": 3.7620866298675537, "max_param_abs": 1.3625353574752808, "policy_loss": -0.0005060401394132441, "value_loss": 3.702852189540863}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:40.587174+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983320.587185, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.59147893541942, "elapsed_seconds": 5729.3241675099125, "env_steps_per_second": 10472.43937430709, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.53049297847314, "eta_seconds": 1909.7747225033042, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 2.1023724642160735e-05, "chaser_grad_norm": 0.14584440737962723, "chaser_new_fields": 82.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.360001087188721, "chaser_subspace_norm": 89.08210754394531, "collisions_per_episode": 1.875, "entropy": 0.046780092641711235, "explorer_grad_norm": 1.2217569127678871, "explorer_new_fields": 66.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 48.625, "explorer_subspace_norm": 28.456058502197266, "final_distance": 4.351083278656006, "max_param_abs": 1.3625353574752808, "policy_loss": -0.0006259714539069705, "value_loss": 3.744202136993408}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:47.055774+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983507.0557888, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6447962673636114, "elapsed_seconds": 5921.266562509001, "env_steps_per_second": 10470.732797702138, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4775214969765324, "eta_seconds": 1719.0773891155166, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004148627770864388, "chaser_grad_norm": 0.22003306448459625, "chaser_new_fields": 82.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 93.22869110107422, "collisions_per_episode": 1.975000023841858, "entropy": 0.09657987207174301, "explorer_grad_norm": 0.7332020699977875, "explorer_new_fields": 60.650001525878906, "explorer_partner_vision": 0.0, "explorer_return": 40.26250076293945, "explorer_subspace_norm": 29.83602523803711, "final_distance": 4.362370491027832, "max_param_abs": 1.3625353574752808, "policy_loss": -0.0010379149946389, "value_loss": 3.9288122057914734}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:14:58.998169+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983698.9981828, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6989014797305573, "elapsed_seconds": 6116.045327030006, "env_steps_per_second": 10464.278235014135, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4247253699326393, "eta_seconds": 1529.0113317575015, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00015387976204925735, "chaser_grad_norm": 0.036796169355511665, "chaser_new_fields": 85.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.930001258850098, "chaser_subspace_norm": 93.2822265625, "collisions_per_episode": 1.899999976158142, "entropy": 0.036916003562510014, "explorer_grad_norm": 0.6658029705286026, "explorer_new_fields": 66.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 49.9375, "explorer_subspace_norm": 34.375267028808594, "final_distance": 4.736596584320068, "max_param_abs": 1.3625342845916748, "policy_loss": -0.0005685174717768859, "value_loss": 3.416408121585846}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:13.776934+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983893.7769487, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7531132384705577, "elapsed_seconds": 6311.207658494008, "env_steps_per_second": 10457.5864987065, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.37187250513011827, "eta_seconds": 1338.7410184684259, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00035642942722022286, "chaser_grad_norm": 0.0578753836452961, "chaser_new_fields": 84.875, "chaser_partner_vision": 0.0, "chaser_return": 6.875000953674316, "chaser_subspace_norm": 90.97238159179688, "collisions_per_episode": 1.774999976158142, "entropy": 0.05746053159236908, "explorer_grad_norm": 1.2020477950572968, "explorer_new_fields": 67.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 49.900001525878906, "explorer_subspace_norm": 35.93645095825195, "final_distance": 4.1623616218566895, "max_param_abs": 1.3880411386489868, "policy_loss": -0.0008748149613388989, "value_loss": 3.580416440963745}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:28.939264+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984088.939273, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8058864161674864, "elapsed_seconds": 6501.191098202951, "env_steps_per_second": 10459.621778968542, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3186858381472035, "eta_seconds": 1147.2690173299327, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00015200773048535364, "chaser_grad_norm": 0.10241780988872051, "chaser_new_fields": 85.625, "chaser_partner_vision": 0.0, "chaser_return": 7.015000820159912, "chaser_subspace_norm": 111.14366912841797, "collisions_per_episode": 1.625, "entropy": 0.04599382635205984, "explorer_grad_norm": 0.43041640147566795, "explorer_new_fields": 75.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 61.78750228881836, "explorer_subspace_norm": 35.29814910888672, "final_distance": 4.566906929016113, "max_param_abs": 1.5263309478759766, "policy_loss": -0.000867961902169867, "value_loss": 3.38794606924057}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:38.922705+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984278.9227195, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8578091695863663, "elapsed_seconds": 6688.113010510919, "env_steps_per_second": 10466.330322168487, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2654013099409095, "eta_seconds": 955.4447157872742, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0001667156830659905, "chaser_grad_norm": 0.11737602576613426, "chaser_new_fields": 86.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.165000915527344, "chaser_subspace_norm": 115.51607513427734, "collisions_per_episode": 2.049999952316284, "entropy": 0.05606973823159933, "explorer_grad_norm": 0.9821602553129196, "explorer_new_fields": 68.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 51.28750228881836, "explorer_subspace_norm": 36.568145751953125, "final_distance": 4.230906009674072, "max_param_abs": 1.579803705215454, "policy_loss": -0.0005939197750421954, "value_loss": 3.638978362083435}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:45.844616+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984465.844627, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9096567948422227, "elapsed_seconds": 6874.764461432002, "env_steps_per_second": 10473.086082283397, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2121840883158025, "eta_seconds": 763.862717936889, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00030691357619508017, "chaser_grad_norm": 0.058673445135354996, "chaser_new_fields": 85.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.880000591278076, "chaser_subspace_norm": 114.53105163574219, "collisions_per_episode": 2.674999952316284, "entropy": 0.0661310013383627, "explorer_grad_norm": 1.010777860879898, "explorer_new_fields": 70.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 54.51250076293945, "explorer_subspace_norm": 39.16767120361328, "final_distance": 4.541614055633545, "max_param_abs": 1.5649555921554565, "policy_loss": -0.0009136478378213297, "value_loss": 3.7951783537864685}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:30:52.496067+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984652.4960825, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9634767002422207, "elapsed_seconds": 7068.516120871995, "env_steps_per_second": 10468.95822752557, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15920081353315305, "eta_seconds": 573.122928719351, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0007817034645435234, "chaser_grad_norm": 0.04898989945650101, "chaser_new_fields": 87.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.42000150680542, "chaser_subspace_norm": 106.81991577148438, "collisions_per_episode": 2.0, "entropy": 0.06963456608355045, "explorer_grad_norm": 0.3882364556193352, "explorer_new_fields": 73.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 58.86249923706055, "explorer_subspace_norm": 36.632198333740234, "final_distance": 3.7910141944885254, "max_param_abs": 1.5649230480194092, "policy_loss": -0.0014278990011655779, "value_loss": 3.5714478492736816}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:06.247727+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984846.2477381, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0169405382011125, "elapsed_seconds": 7260.985937524005, "env_steps_per_second": 10466.898111899662, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10615476516847959, "eta_seconds": 382.15715460652655, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005022426746123382, "chaser_grad_norm": 0.040396939031779766, "chaser_new_fields": 87.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.315001010894775, "chaser_subspace_norm": 99.07245635986328, "collisions_per_episode": 2.1500000953674316, "entropy": 0.04066335689276457, "explorer_grad_norm": 1.7732303738594055, "explorer_new_fields": 73.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 60.212501525878906, "explorer_subspace_norm": 35.64181900024414, "final_distance": 4.642685413360596, "max_param_abs": 1.6103510856628418, "policy_loss": -0.00040899771487090675, "value_loss": 3.4562589526176453}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:18.717544+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985038.7175567, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.069375607725265, "elapsed_seconds": 7449.752187810955, "env_steps_per_second": 10470.146930205423, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05306091301859655, "eta_seconds": 191.01928686694757, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0006225803429906335, "chaser_grad_norm": 0.04072236688807607, "chaser_new_fields": 87.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 7.375000953674316, "chaser_subspace_norm": 109.83747100830078, "collisions_per_episode": 1.899999976158142, "entropy": 0.05799243692308664, "explorer_grad_norm": 1.7096837162971497, "explorer_new_fields": 75.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 35.779136657714844, "final_distance": 3.7842938899993896, "max_param_abs": 1.6103510856628418, "policy_loss": -0.0008497528235891227, "value_loss": 3.4987517595291138}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:27.483793+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985227.4838045, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.119818290252491, "elapsed_seconds": 7631.345844908967, "env_steps_per_second": 10483.078820673512, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:29.077450+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985409.0774622, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.119901634450265, "elapsed_seconds": 7631.6458840209525, "env_steps_per_second": 10482.666677119156, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:29.377490+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985409.377503, "update": 20000} diff --git a/non_social/seed_0003/attempt_01/status.json b/non_social/seed_0003/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..31d264e0d747d201d2f462a1bf8202e46102a7e8 --- /dev/null +++ b/non_social/seed_0003/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.119901634450265, + "elapsed_seconds": 7631.6458840209525, + "env_steps_per_second": 10482.666677119156, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0002841827333717878, + "chaser_grad_norm": 0.04017048981040716, + "chaser_new_fields": 87.07500457763672, + "chaser_partner_vision": 0.0, + "chaser_return": 7.300001621246338, + "chaser_subspace_norm": 112.96394348144531, + "collisions_per_episode": 1.725000023841858, + "entropy": 0.024153232108801603, + "explorer_grad_norm": 0.8466866165399551, + "explorer_new_fields": 75.2249984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 62.125, + "explorer_subspace_norm": 36.111541748046875, + "final_distance": 4.144174098968506, + "max_param_abs": 1.6294050216674805, + "policy_loss": -0.0005663236785924397, + "value_loss": 3.4905534386634827 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:29.377490+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985409.377503, + "update": 20000 +} diff --git a/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c13d6522b216aa3859af5ce0fdee1100936fb4ce --- /dev/null +++ b/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1fa11ee31375fa76d13192833c9376692c520fcc925fee9c6b76a18abd4318fa +size 11460392 diff --git a/non_social/seed_0004/attempt_01/config.json b/non_social/seed_0004/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..08a2c583501d053993c744f2b0b44cc594306537 --- /dev/null +++ b/non_social/seed_0004/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", + "seed": 4, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0004/attempt_01/metrics.jsonl b/non_social/seed_0004/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..00ecaeaa76663afe997a5743cf6217da14275add --- /dev/null +++ b/non_social/seed_0004/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005172178980622751, "elapsed_seconds": 1.8619844330241904, "env_steps_per_second": 2148.245672227934, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.34384074334744, "eta_seconds": 37237.82667605078, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -2.8136250477477087e-05, "chaser_grad_norm": 0.3089674226939678, "chaser_new_fields": 36.75, "chaser_partner_vision": 0.0, "chaser_return": -2.695000171661377, "chaser_subspace_norm": 0.0019733882509171963, "collisions_per_episode": 1.524999976158142, "entropy": 2.772531807422638, "explorer_grad_norm": 0.10119399800896645, "explorer_new_fields": 35.0, "explorer_partner_vision": 0.0, "explorer_return": 2.125, "explorer_subspace_norm": 0.0021857190877199173, "final_distance": 5.1569952964782715, "max_param_abs": 0.15428043901920319, "policy_loss": -0.0005164051181640872, "value_loss": 3.1183584928512573}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.512933+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.5129604, "update": 1} +{"device": "cuda", "elapsed_hours": 0.053556909304170404, "elapsed_seconds": 192.80487349501345, "env_steps_per_second": 10373.18177567605, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.088719462862646, "eta_seconds": 7519.390066305526, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0007860452005472585, "chaser_grad_norm": 0.08322609402239323, "chaser_new_fields": 73.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.5400004386901855, "chaser_subspace_norm": 13.988959312438965, "collisions_per_episode": 2.799999952316284, "entropy": 0.5719539374113083, "explorer_grad_norm": 0.5221506729722023, "explorer_new_fields": 63.22500228881836, "explorer_partner_vision": 0.0, "explorer_return": 44.3125, "explorer_subspace_norm": 6.704671382904053, "final_distance": 5.773620128631592, "max_param_abs": 0.2983703017234802, "policy_loss": -0.002123032832052507, "value_loss": 4.772855520248413}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:30.455813+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977970.4558268, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10610932520558385, "elapsed_seconds": 381.99357074010186, "env_steps_per_second": 10471.380427293872, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.016077178906093, "eta_seconds": 7257.877844061935, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0003023507916948276, "chaser_grad_norm": 0.1278394926339388, "chaser_new_fields": 73.875, "chaser_partner_vision": 0.0, "chaser_return": 4.640000820159912, "chaser_subspace_norm": 25.73876953125, "collisions_per_episode": 3.125, "entropy": 0.42118658125400543, "explorer_grad_norm": 0.8685867190361023, "explorer_new_fields": 64.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 46.60000228881836, "explorer_subspace_norm": 7.97187614440918, "final_distance": 5.476497173309326, "max_param_abs": 0.3355414867401123, "policy_loss": -0.0012387437326850659, "value_loss": 4.438918590545654}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:39.644509+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978159.6445227, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15936692130722804, "elapsed_seconds": 573.720916706021, "env_steps_per_second": 10458.046456539507, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9655253627891458, "eta_seconds": 7075.891306040925, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00047077342389334476, "chaser_grad_norm": 0.1952325776219368, "chaser_new_fields": 79.375, "chaser_partner_vision": 0.0, "chaser_return": 5.7350006103515625, "chaser_subspace_norm": 38.00165939331055, "collisions_per_episode": 2.674999952316284, "entropy": 0.27293500304222107, "explorer_grad_norm": 0.8026920929551125, "explorer_new_fields": 68.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 51.51250076293945, "explorer_subspace_norm": 10.24711799621582, "final_distance": 5.400847911834717, "max_param_abs": 0.3939908444881439, "policy_loss": -0.0014115015142603937, "value_loss": 4.170977592468262}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:51.371855+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978351.3718674, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21294707654027864, "elapsed_seconds": 766.6094755450031, "env_steps_per_second": 10435.561071447215, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9165236888625077, "eta_seconds": 6899.485279905028, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 5.9287332524582403e-05, "chaser_grad_norm": 0.13057871628552675, "chaser_new_fields": 79.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.7350006103515625, "chaser_subspace_norm": 44.125667572021484, "collisions_per_episode": 2.625, "entropy": 0.2455620989203453, "explorer_grad_norm": 0.8171755969524384, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 11.008387565612793, "final_distance": 4.741165637969971, "max_param_abs": 0.415237694978714, "policy_loss": -0.001017891316784869, "value_loss": 4.246682167053223}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:04.260413+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978544.2604258, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26613591383974483, "elapsed_seconds": 958.0892898230813, "env_steps_per_second": 10437.440545699637, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8629513968782137, "eta_seconds": 6706.625028761569, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00040906606965585013, "chaser_grad_norm": 0.11256977915763855, "chaser_new_fields": 77.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.455000877380371, "chaser_subspace_norm": 48.35102462768555, "collisions_per_episode": 2.0250000953674316, "entropy": 0.21318147704005241, "explorer_grad_norm": 0.4030519798398018, "explorer_new_fields": 65.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 46.900001525878906, "explorer_subspace_norm": 12.477490425109863, "final_distance": 5.532765865325928, "max_param_abs": 0.4690392017364502, "policy_loss": -0.001173720835936365, "value_loss": 4.0489890575408936}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:15.740226+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978735.7402377, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.32004238563667364, "elapsed_seconds": 1152.152588292025, "env_steps_per_second": 10415.28710861905, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8135735186078172, "eta_seconds": 6528.864666988142, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.000210464247233616, "chaser_grad_norm": 0.2844482623040676, "chaser_new_fields": 78.25, "chaser_partner_vision": 0.0, "chaser_return": 5.510000705718994, "chaser_subspace_norm": 49.22774124145508, "collisions_per_episode": 3.2250001430511475, "entropy": 0.18961040675640106, "explorer_grad_norm": 0.7613688707351685, "explorer_new_fields": 65.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 47.57500076293945, "explorer_subspace_norm": 13.477876663208008, "final_distance": 5.25750732421875, "max_param_abs": 0.4974443316459656, "policy_loss": -0.00167304977029481, "value_loss": 4.019002318382263}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:29.803526+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978929.8035398, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.37374965273280397, "elapsed_seconds": 1345.4987498380942, "env_steps_per_second": 10405.06355110671, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7619626485975042, "eta_seconds": 6343.065534951015, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00028538182759058756, "chaser_grad_norm": 0.07672451809048653, "chaser_new_fields": 81.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.240001201629639, "chaser_subspace_norm": 57.149295806884766, "collisions_per_episode": 1.75, "entropy": 0.14440559968352318, "explorer_grad_norm": 0.4887581095099449, "explorer_new_fields": 67.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 51.28750228881836, "explorer_subspace_norm": 14.017056465148926, "final_distance": 5.07213830947876, "max_param_abs": 0.5375002026557922, "policy_loss": -0.0006884204031649688, "value_loss": 3.8559442162513733}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:43.149688+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979123.149702, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.42603960544362457, "elapsed_seconds": 1533.7425795970485, "env_steps_per_second": 10431.998311087895, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7041584217744983, "eta_seconds": 6134.970318388194, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 2.289587441229972e-05, "chaser_grad_norm": 0.13343936763703823, "chaser_new_fields": 80.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.935000896453857, "chaser_subspace_norm": 61.63414001464844, "collisions_per_episode": 1.8000000715255737, "entropy": 0.1238627340644598, "explorer_grad_norm": 2.2425120174884796, "explorer_new_fields": 69.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 53.16250228881836, "explorer_subspace_norm": 16.035688400268555, "final_distance": 4.023276329040527, "max_param_abs": 0.5641061663627625, "policy_loss": -0.0004263466441152275, "value_loss": 3.7302972078323364}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:51.393517+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979311.3935325, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.47904960297614, "elapsed_seconds": 1724.578570714104, "env_steps_per_second": 10437.33252034244, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6500597435844822, "eta_seconds": 5940.2150769041355, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0004124707017745255, "chaser_grad_norm": 0.08845847379416227, "chaser_new_fields": 80.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.070001125335693, "chaser_subspace_norm": 63.17826843261719, "collisions_per_episode": 1.875, "entropy": 0.1098443791270256, "explorer_grad_norm": 1.544716328382492, "explorer_new_fields": 70.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 54.400001525878906, "explorer_subspace_norm": 17.435523986816406, "final_distance": 5.642947673797607, "max_param_abs": 0.5757636427879333, "policy_loss": -0.0009579531205829994, "value_loss": 3.6687695384025574}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:05:02.229509+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979502.2295232, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5323221233866773, "elapsed_seconds": 1916.359644192038, "env_steps_per_second": 10436.454378808554, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.596966370160032, "eta_seconds": 5749.078932576115, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005175808601550203, "chaser_grad_norm": 0.042110917158424854, "chaser_new_fields": 80.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.995000839233398, "chaser_subspace_norm": 63.17192077636719, "collisions_per_episode": 2.049999952316284, "entropy": 0.07635793276131153, "explorer_grad_norm": 0.7532302737236023, "explorer_new_fields": 71.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 56.16250228881836, "explorer_subspace_norm": 17.782289505004883, "final_distance": 4.473559856414795, "max_param_abs": 0.5924593806266785, "policy_loss": -0.0010896424569608598, "value_loss": 3.6565895080566406}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:14.010583+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979694.0105991, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5851372999564046, "elapsed_seconds": 2106.4942798430566, "env_steps_per_second": 10443.89258993815, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5426346998850669, "eta_seconds": 5553.484919586241, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00044180230212453875, "chaser_grad_norm": 0.1212445329874754, "chaser_new_fields": 81.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.245001316070557, "chaser_subspace_norm": 58.884822845458984, "collisions_per_episode": 2.325000047683716, "entropy": 0.09421484544873238, "explorer_grad_norm": 1.5021811425685883, "explorer_new_fields": 72.0, "explorer_partner_vision": 0.0, "explorer_return": 57.17499923706055, "explorer_subspace_norm": 18.92897605895996, "final_distance": 4.8955488204956055, "max_param_abs": 0.6221098899841309, "policy_loss": -0.0009583394189789374, "value_loss": 3.611276686191559}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:24.145219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979884.1452312, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6387129269636352, "elapsed_seconds": 2299.366537069087, "env_steps_per_second": 10437.657334350819, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4903301629151489, "eta_seconds": 5365.188586494536, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00039409020403535777, "chaser_grad_norm": 0.09305899403989315, "chaser_new_fields": 82.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.380001068115234, "chaser_subspace_norm": 64.45477294921875, "collisions_per_episode": 1.8000000715255737, "entropy": 0.06290358863770962, "explorer_grad_norm": 0.7457608133554459, "explorer_new_fields": 72.0, "explorer_partner_vision": 0.0, "explorer_return": 57.36249923706055, "explorer_subspace_norm": 20.856107711791992, "final_distance": 4.556450843811035, "max_param_abs": 0.6583436131477356, "policy_loss": -0.0010948784189501737, "value_loss": 3.646681547164917}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:37.017476+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980077.0174906, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6917934254572417, "elapsed_seconds": 2490.45633164607, "env_steps_per_second": 10439.85380093586, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4368017297958096, "eta_seconds": 5172.486227264914, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": -2.5081788404685e-05, "chaser_grad_norm": 0.12715299986302853, "chaser_new_fields": 79.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.805001258850098, "chaser_subspace_norm": 72.15359497070312, "collisions_per_episode": 2.4750001430511475, "entropy": 0.043966300785541534, "explorer_grad_norm": 1.158193051815033, "explorer_new_fields": 72.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.70000076293945, "explorer_subspace_norm": 22.238468170166016, "final_distance": 5.062656402587891, "max_param_abs": 0.6664208173751831, "policy_loss": -0.0011894201640965107, "value_loss": 3.6388476490974426}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:48.107268+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980268.1072805, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7461092098250002, "elapsed_seconds": 2685.9931553700007, "env_steps_per_second": 10424.449497952255, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3856313896750005, "eta_seconds": 4988.273002830002, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -9.518431942340255e-05, "chaser_grad_norm": 0.24729258939623833, "chaser_new_fields": 82.0, "chaser_partner_vision": 0.0, "chaser_return": 6.27000093460083, "chaser_subspace_norm": 72.15290069580078, "collisions_per_episode": 1.9250000715255737, "entropy": 0.05438780225813389, "explorer_grad_norm": 0.539922721683979, "explorer_new_fields": 72.375, "explorer_partner_vision": 0.0, "explorer_return": 57.73749923706055, "explorer_subspace_norm": 22.831787109375, "final_distance": 4.5554423332214355, "max_param_abs": 0.690929651260376, "policy_loss": -0.0003015694060017893, "value_loss": 3.6087552309036255}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:21:03.644093+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980463.6441069, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7988193393944595, "elapsed_seconds": 2875.749621820054, "env_steps_per_second": 10432.062573310219, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3313655656574321, "eta_seconds": 4792.916036366756, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 2.0460200565980413e-05, "chaser_grad_norm": 0.046512825414538383, "chaser_new_fields": 78.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.670001029968262, "chaser_subspace_norm": 76.57718658447266, "collisions_per_episode": 2.0, "entropy": 0.07929566130042076, "explorer_grad_norm": 0.574237510561943, "explorer_new_fields": 71.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 56.01250076293945, "explorer_subspace_norm": 24.11458969116211, "final_distance": 5.034781455993652, "max_param_abs": 0.7641788721084595, "policy_loss": -0.000537539648234997, "value_loss": 3.6687775254249573}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:13.400559+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980653.4005706, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8523993392141872, "elapsed_seconds": 3068.637621171074, "env_steps_per_second": 10428.080454735462, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2785990088212809, "eta_seconds": 4602.956431756611, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0003013324536510764, "chaser_grad_norm": 0.10626604780554771, "chaser_new_fields": 80.875, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 75.4118881225586, "collisions_per_episode": 2.299999952316284, "entropy": 0.04415551479905844, "explorer_grad_norm": 0.380926713347435, "explorer_new_fields": 70.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 55.375, "explorer_subspace_norm": 24.33274269104004, "final_distance": 5.808345794677734, "max_param_abs": 0.764813244342804, "policy_loss": -0.0009845115933551085, "value_loss": 3.601925015449524}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:26.288559+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980846.2885742, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9060146509508357, "elapsed_seconds": 3261.6527434230084, "env_steps_per_second": 10424.163046958212, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.225784527757013, "eta_seconds": 4412.824299925247, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 6.580311044329656e-05, "chaser_grad_norm": 0.1789839193224907, "chaser_new_fields": 75.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.990001201629639, "chaser_subspace_norm": 74.4191665649414, "collisions_per_episode": 4.599999904632568, "entropy": 0.08674476854503155, "explorer_grad_norm": 1.1754572913050652, "explorer_new_fields": 67.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 50.16250228881836, "explorer_subspace_norm": 26.081832885742188, "final_distance": 4.7001051902771, "max_param_abs": 0.8178247809410095, "policy_loss": -0.0007025698686478066, "value_loss": 3.8433682322502136}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:39.303680+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981039.3036907, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9580587746766913, "elapsed_seconds": 3449.011588836089, "env_steps_per_second": 10437.772988796665, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.170960724604845, "eta_seconds": 4215.458608577442, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00016174525336021955, "chaser_grad_norm": 0.06815970968455076, "chaser_new_fields": 79.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.785000801086426, "chaser_subspace_norm": 76.56777954101562, "collisions_per_episode": 2.700000047683716, "entropy": 0.044688607566058636, "explorer_grad_norm": 0.2295846790075302, "explorer_new_fields": 70.5, "explorer_partner_vision": 0.0, "explorer_return": 54.8125, "explorer_subspace_norm": 26.114025115966797, "final_distance": 4.678368091583252, "max_param_abs": 0.8783822059631348, "policy_loss": -0.0003650099185179556, "value_loss": 3.5541728734970093}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:46.662527+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981226.6625412, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0098573817752832, "elapsed_seconds": 3635.4865743910195, "env_steps_per_second": 10452.5210649046, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1161581588042604, "eta_seconds": 4018.1693716953378, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00032732175971650346, "chaser_grad_norm": 0.0916148442775011, "chaser_new_fields": 78.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.525001049041748, "chaser_subspace_norm": 78.609375, "collisions_per_episode": 2.125, "entropy": 0.09363203309476376, "explorer_grad_norm": 0.9501346349716187, "explorer_new_fields": 69.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 53.35000228881836, "explorer_subspace_norm": 25.70167350769043, "final_distance": 5.711217403411865, "max_param_abs": 0.9460973739624023, "policy_loss": -0.000831880499201354, "value_loss": 3.7667624950408936}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:53.137511+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981413.1375234, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0617287772019497, "elapsed_seconds": 3822.223597927019, "env_steps_per_second": 10465.11251243752, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0617287772019497, "eta_seconds": 3822.223597927019, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0007615951672133303, "chaser_grad_norm": 0.0533189382404089, "chaser_new_fields": 77.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.390000820159912, "chaser_subspace_norm": 74.69845581054688, "collisions_per_episode": 2.299999952316284, "entropy": 0.06492173299193382, "explorer_grad_norm": 0.9686664491891861, "explorer_new_fields": 72.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 58.375, "explorer_subspace_norm": 25.46718406677246, "final_distance": 5.02777624130249, "max_param_abs": 0.9460973739624023, "policy_loss": -0.0006033623203336447, "value_loss": 3.573467791080475}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:39:59.874536+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981599.87455, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1143190825119704, "elapsed_seconds": 4011.548697043094, "env_steps_per_second": 10469.771943927324, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0081934556060685, "eta_seconds": 3629.496440181847, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00032360340482752437, "chaser_grad_norm": 0.12506836093962193, "chaser_new_fields": 79.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.810000896453857, "chaser_subspace_norm": 77.67725372314453, "collisions_per_episode": 1.274999976158142, "entropy": 0.055369116365909576, "explorer_grad_norm": 0.9641833901405334, "explorer_new_fields": 72.75, "explorer_partner_vision": 0.0, "explorer_return": 58.525001525878906, "explorer_subspace_norm": 26.72629737854004, "final_distance": 5.076308727264404, "max_param_abs": 0.9460973739624023, "policy_loss": -0.0006941037323571209, "value_loss": 3.6052610874176025}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:09.199633+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981789.1996443, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1676901565819733, "elapsed_seconds": 4203.684563695104, "env_steps_per_second": 10467.008010068985, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9553828553852509, "eta_seconds": 3439.378279386903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": -0.00016675485911454846, "chaser_grad_norm": 0.062450320459902287, "chaser_new_fields": 79.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.720000743865967, "chaser_subspace_norm": 80.6820068359375, "collisions_per_episode": 2.2250001430511475, "entropy": 0.05668195057660341, "explorer_grad_norm": 0.5559889897704124, "explorer_new_fields": 73.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 59.53750228881836, "explorer_subspace_norm": 25.628742218017578, "final_distance": 5.253024578094482, "max_param_abs": 1.0006669759750366, "policy_loss": -0.0008007144025759805, "value_loss": 3.53891921043396}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:21.335501+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981981.335516, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2211785787213982, "elapsed_seconds": 4396.242883397033, "env_steps_per_second": 10463.480117016466, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9026102538375552, "eta_seconds": 3249.396913815199, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 9.018279330574863e-05, "chaser_grad_norm": 0.101340027526021, "chaser_new_fields": 78.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.560000896453857, "chaser_subspace_norm": 77.86398315429688, "collisions_per_episode": 1.875, "entropy": 0.06060435064136982, "explorer_grad_norm": 0.7840324714779854, "explorer_new_fields": 71.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.025001525878906, "explorer_subspace_norm": 26.74220085144043, "final_distance": 5.201775550842285, "max_param_abs": 1.0118290185928345, "policy_loss": -0.0009246987847149057, "value_loss": 3.630063235759735}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:33.893820+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982173.8938322, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.275133201693631, "elapsed_seconds": 4590.479526097071, "env_steps_per_second": 10456.423937220057, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8500888011290872, "eta_seconds": 3060.3196840647142, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005607759615676722, "chaser_grad_norm": 0.044707116205245256, "chaser_new_fields": 75.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.085000514984131, "chaser_subspace_norm": 76.03313446044922, "collisions_per_episode": 2.075000047683716, "entropy": 0.041923641227185726, "explorer_grad_norm": 0.551770307123661, "explorer_new_fields": 71.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 55.75, "explorer_subspace_norm": 28.322118759155273, "final_distance": 5.025094032287598, "max_param_abs": 1.027726173400879, "policy_loss": -0.001191765245517118, "value_loss": 3.7599114179611206}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:48.130464+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982368.130477, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3281739703066866, "elapsed_seconds": 4781.426293104072, "env_steps_per_second": 10457.130767049912, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.796904382184012, "eta_seconds": 2868.8557758624434, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0004973404827692818, "chaser_grad_norm": 0.07222611829638481, "chaser_new_fields": 80.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.0550007820129395, "chaser_subspace_norm": 82.09326934814453, "collisions_per_episode": 1.9500000476837158, "entropy": 0.06519300304353237, "explorer_grad_norm": 0.6737918704748154, "explorer_new_fields": 71.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 56.462501525878906, "explorer_subspace_norm": 27.60597038269043, "final_distance": 5.678858280181885, "max_param_abs": 1.0354223251342773, "policy_loss": -0.0009853024423360779, "value_loss": 3.520586371421814}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:59.077232+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982559.0772452, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3820146839277974, "elapsed_seconds": 4975.252862140071, "env_steps_per_second": 10451.730081038044, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7441617528841986, "eta_seconds": 2678.982310383115, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0009375374997868113, "chaser_grad_norm": 0.19851895049214363, "chaser_new_fields": 77.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.330001354217529, "chaser_subspace_norm": 83.45472717285156, "collisions_per_episode": 2.1000001430511475, "entropy": 0.09524623304605484, "explorer_grad_norm": 1.398541808128357, "explorer_new_fields": 70.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 54.92499923706055, "explorer_subspace_norm": 25.997703552246094, "final_distance": 5.215616703033447, "max_param_abs": 1.0785167217254639, "policy_loss": -0.0008824701986718253, "value_loss": 3.856283962726593}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:12.903800+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982752.9038143, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4347511428583595, "elapsed_seconds": 5165.104114290094, "env_steps_per_second": 10454.774735440527, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6908061058206916, "eta_seconds": 2486.9019809544898, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 2.0289660930754294e-05, "chaser_grad_norm": 0.23099471256136894, "chaser_new_fields": 77.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.465000629425049, "chaser_subspace_norm": 81.31656646728516, "collisions_per_episode": 2.2250001430511475, "entropy": 0.07878018915653229, "explorer_grad_norm": 1.7294500023126602, "explorer_new_fields": 73.625, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 28.85605239868164, "final_distance": 4.860409736633301, "max_param_abs": 1.0976755619049072, "policy_loss": -0.0007817273482215725, "value_loss": 3.5737887620925903}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:22.755051+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982942.755062, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.486558299825863, "elapsed_seconds": 5351.609879373107, "env_steps_per_second": 10464.14093371094, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6370964142110842, "eta_seconds": 2293.547091159903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0003763100500743799, "chaser_grad_norm": 0.10625240951776505, "chaser_new_fields": 79.5, "chaser_partner_vision": 0.0, "chaser_return": 5.820001125335693, "chaser_subspace_norm": 79.58535766601562, "collisions_per_episode": 1.899999976158142, "entropy": 0.053851254284381866, "explorer_grad_norm": 0.965865969657898, "explorer_new_fields": 73.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 60.4375, "explorer_subspace_norm": 29.14771842956543, "final_distance": 4.30214262008667, "max_param_abs": 1.1747386455535889, "policy_loss": -0.0010073953876279518, "value_loss": 3.546108365058899}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:29.260817+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983129.26083, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5396230980564076, "elapsed_seconds": 5542.643153003068, "env_steps_per_second": 10464.321515732965, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5839949682282926, "eta_seconds": 2102.381885621853, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -4.391738387246949e-05, "chaser_grad_norm": 0.15401971153914928, "chaser_new_fields": 77.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.385000705718994, "chaser_subspace_norm": 89.02958679199219, "collisions_per_episode": 2.2750000953674316, "entropy": 0.09169591963291168, "explorer_grad_norm": 1.5349881052970886, "explorer_new_fields": 73.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 28.978702545166016, "final_distance": 5.557142734527588, "max_param_abs": 1.2972217798233032, "policy_loss": -0.00032433308058532617, "value_loss": 3.604218602180481}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:40.294092+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983320.2941046, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5931413366728036, "elapsed_seconds": 5735.308812022093, "env_steps_per_second": 10461.511658139618, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5310471122242678, "eta_seconds": 1911.769604007364, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00012728126651757096, "chaser_grad_norm": 0.08748726733028889, "chaser_new_fields": 80.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.000000953674316, "chaser_subspace_norm": 91.54847717285156, "collisions_per_episode": 1.850000023841858, "entropy": 0.04745035618543625, "explorer_grad_norm": 1.3337886780500412, "explorer_new_fields": 72.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 58.1875, "explorer_subspace_norm": 29.716659545898438, "final_distance": 4.856721878051758, "max_param_abs": 1.3316469192504883, "policy_loss": -0.000506102030573885, "value_loss": 3.4923959374427795}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:52.959749+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983512.959762, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6466134691864136, "elapsed_seconds": 5927.808489071089, "env_steps_per_second": 10459.17730208515, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47804907169928135, "eta_seconds": 1720.9766581174129, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0002815115874464369, "chaser_grad_norm": 0.08174392022192478, "chaser_new_fields": 81.75, "chaser_partner_vision": 0.0, "chaser_return": 6.200000762939453, "chaser_subspace_norm": 81.99629974365234, "collisions_per_episode": 1.9250000715255737, "entropy": 0.07287757471203804, "explorer_grad_norm": 1.5110264867544174, "explorer_new_fields": 72.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.8125, "explorer_subspace_norm": 28.256479263305664, "final_distance": 5.5624613761901855, "max_param_abs": 1.3296668529510498, "policy_loss": -0.0005825825121732464, "value_loss": 3.5315281748771667}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:15:05.459428+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983705.4594398, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6990556171778008, "elapsed_seconds": 6116.600221840083, "env_steps_per_second": 10463.328921102287, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4247639042944502, "eta_seconds": 1529.1500554600207, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00035150080059431765, "chaser_grad_norm": 0.06088258605450392, "chaser_new_fields": 81.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.210000991821289, "chaser_subspace_norm": 81.1490707397461, "collisions_per_episode": 2.0, "entropy": 0.06425238773226738, "explorer_grad_norm": 1.225659355521202, "explorer_new_fields": 72.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.51250076293945, "explorer_subspace_norm": 28.04905128479004, "final_distance": 4.756880283355713, "max_param_abs": 1.3296668529510498, "policy_loss": -0.0008116592236362408, "value_loss": 3.5536530017852783}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:14.251159+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983894.2511725, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7529047566714062, "elapsed_seconds": 6310.457124017063, "env_steps_per_second": 10458.830272185769, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.37182828171817706, "eta_seconds": 1338.5818141854375, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0002101763966899095, "chaser_grad_norm": 0.08745403215289116, "chaser_new_fields": 80.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.915000915527344, "chaser_subspace_norm": 77.29298400878906, "collisions_per_episode": 2.0250000953674316, "entropy": 0.05143814906477928, "explorer_grad_norm": 2.0979818552732468, "explorer_new_fields": 73.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 59.837501525878906, "explorer_subspace_norm": 29.637107849121094, "final_distance": 4.726080417633057, "max_param_abs": 1.3296668529510498, "policy_loss": -0.0008582087642281522, "value_loss": 3.4864986538887024}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:28.108061+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984088.1080728, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.806505419253889, "elapsed_seconds": 6503.419509314001, "env_steps_per_second": 10456.03776638005, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31879507398598045, "eta_seconds": 1147.6622663495295, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00012691038279260408, "chaser_grad_norm": 0.041202503722161055, "chaser_new_fields": 79.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.770001411437988, "chaser_subspace_norm": 77.6485824584961, "collisions_per_episode": 1.600000023841858, "entropy": 0.08899831399321556, "explorer_grad_norm": 1.0570008531212807, "explorer_new_fields": 72.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.8125, "explorer_subspace_norm": 27.71615982055664, "final_distance": 5.0411858558654785, "max_param_abs": 1.3296668529510498, "policy_loss": -0.0010536232222833064, "value_loss": 3.596246838569641}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:41.070447+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984281.0704563, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8586597965530625, "elapsed_seconds": 6691.175267591025, "env_steps_per_second": 10461.540342404092, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2655228280790089, "eta_seconds": 955.8821810844321, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00025314518904195893, "chaser_grad_norm": 0.07486117631196976, "chaser_new_fields": 80.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.045001029968262, "chaser_subspace_norm": 85.50122833251953, "collisions_per_episode": 1.5500000715255737, "entropy": 0.05326128378510475, "explorer_grad_norm": 0.27461810410022736, "explorer_new_fields": 75.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.900001525878906, "explorer_subspace_norm": 29.32468605041504, "final_distance": 5.2069315910339355, "max_param_abs": 1.3296668529510498, "policy_loss": -0.0006508207198940852, "value_loss": 3.360384225845337}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:48.826206+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984468.8262203, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.911432221964448, "elapsed_seconds": 6881.1559990720125, "env_steps_per_second": 10463.358192970754, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2123813579960498, "eta_seconds": 764.5728887857792, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0001713470099711234, "chaser_grad_norm": 0.06350186932832003, "chaser_new_fields": 80.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.89500093460083, "chaser_subspace_norm": 84.94845581054688, "collisions_per_episode": 2.3500001430511475, "entropy": 0.0768278893083334, "explorer_grad_norm": 1.4199564009904861, "explorer_new_fields": 76.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 63.17499923706055, "explorer_subspace_norm": 28.288578033447266, "final_distance": 5.2691240310668945, "max_param_abs": 1.3296668529510498, "policy_loss": -0.0011726631184529879, "value_loss": 3.4914387464523315}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:30:58.806938+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984658.806953, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9652394508980797, "elapsed_seconds": 7074.862023233087, "env_steps_per_second": 10459.567940263987, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15934373926200646, "eta_seconds": 573.6374613432232, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00023518021283819301, "chaser_grad_norm": 0.03835074044764042, "chaser_new_fields": 81.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.205001354217529, "chaser_subspace_norm": 87.82594299316406, "collisions_per_episode": 1.600000023841858, "entropy": 0.07425666600465775, "explorer_grad_norm": 0.4650353118777275, "explorer_new_fields": 76.125, "explorer_partner_vision": 0.0, "explorer_return": 63.400001525878906, "explorer_subspace_norm": 28.923202514648438, "final_distance": 5.315260410308838, "max_param_abs": 1.348924994468689, "policy_loss": -0.0007809530487250527, "value_loss": 3.353395998477936}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:12.512962+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984852.5129755, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0191051735064116, "elapsed_seconds": 7268.778624623083, "env_steps_per_second": 10455.676795899246, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10626869334244272, "eta_seconds": 382.5672960327938, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00034193054558584457, "chaser_grad_norm": 0.0797239001840353, "chaser_new_fields": 80.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.935000896453857, "chaser_subspace_norm": 88.2709732055664, "collisions_per_episode": 1.6750000715255737, "entropy": 0.07029056362807751, "explorer_grad_norm": 0.8059673309326172, "explorer_new_fields": 74.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.29999923706055, "explorer_subspace_norm": 29.052474975585938, "final_distance": 5.567540645599365, "max_param_abs": 1.2921407222747803, "policy_loss": -0.0005515422968613848, "value_loss": 3.479243218898773}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:26.429563+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985046.4295802, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.0718003725330667, "elapsed_seconds": 7458.481341119041, "env_steps_per_second": 10457.893025753308, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.053123086475206846, "eta_seconds": 191.24311131074464, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0003529285328892767, "chaser_grad_norm": 0.14526221714913845, "chaser_new_fields": 79.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.835000514984131, "chaser_subspace_norm": 89.80217742919922, "collisions_per_episode": 2.2250001430511475, "entropy": 0.07000009156763554, "explorer_grad_norm": 0.5450664982199669, "explorer_new_fields": 75.625, "explorer_partner_vision": 0.0, "explorer_return": 62.57500076293945, "explorer_subspace_norm": 31.11674690246582, "final_distance": 5.476113796234131, "max_param_abs": 1.2834994792938232, "policy_loss": -0.00036793233290666194, "value_loss": 3.403395414352417}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:36.132277+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985236.1322887, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1224889203575246, "elapsed_seconds": 7640.9601132870885, "env_steps_per_second": 10469.88844515569, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:38.611050+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985418.6110647, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.122494922226114, "elapsed_seconds": 7640.98172001401, "env_steps_per_second": 10469.858839009672, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:38.632658+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985418.632673, "update": 20000} diff --git a/non_social/seed_0004/attempt_01/status.json b/non_social/seed_0004/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..b58dfd4550dd7fff68e8510d3256e267999894b9 --- /dev/null +++ b/non_social/seed_0004/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.122494922226114, + "elapsed_seconds": 7640.98172001401, + "env_steps_per_second": 10469.858839009672, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.00037036755928841064, + "chaser_grad_norm": 0.08058446925133467, + "chaser_new_fields": 82.17500305175781, + "chaser_partner_vision": 0.0, + "chaser_return": 6.330001354217529, + "chaser_subspace_norm": 92.44800567626953, + "collisions_per_episode": 1.8250000476837158, + "entropy": 0.04575740918517113, + "explorer_grad_norm": 0.3279200941324234, + "explorer_new_fields": 77.9749984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 66.0999984741211, + "explorer_subspace_norm": 29.62271499633789, + "final_distance": 5.308696269989014, + "max_param_abs": 1.3726928234100342, + "policy_loss": -0.0006299224167634065, + "value_loss": 3.322889506816864 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:38.632658+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985418.632673, + "update": 20000 +} diff --git a/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..960be5cf02dea2f8ffc8a5c6fcd27d99b348dc09 --- /dev/null +++ b/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5debd7cc50f687b861a1987d4433d6434b53e3ca6cd4c4743ac9192794e9ed0c +size 11460392 diff --git a/non_social/seed_0005/attempt_01/config.json b/non_social/seed_0005/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..f1923206fedaa058f388b7c201d370dfc6e2fc04 --- /dev/null +++ b/non_social/seed_0005/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", + "seed": 5, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0005/attempt_01/metrics.jsonl b/non_social/seed_0005/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..d2e453a8e480ae7d151df7293d3e750dc940ce8b --- /dev/null +++ b/non_social/seed_0005/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0004448346591865023, "elapsed_seconds": 1.6014047730714083, "env_steps_per_second": 2497.806967521531, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 8.89624834907086, "eta_seconds": 32026.494056655094, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 6.800122100691919e-06, "chaser_grad_norm": 0.39571650326251984, "chaser_new_fields": 33.125, "chaser_partner_vision": 0.0, "chaser_return": -3.4250004291534424, "chaser_subspace_norm": 0.002188874641433358, "collisions_per_episode": 1.375, "entropy": 2.7724965810775757, "explorer_grad_norm": 0.08912728354334831, "explorer_new_fields": 34.79999923706055, "explorer_partner_vision": 0.0, "explorer_return": 1.8625000715255737, "explorer_subspace_norm": 0.0027146032080054283, "final_distance": 5.587429046630859, "max_param_abs": 0.15429167449474335, "policy_loss": -0.0005914238751287826, "value_loss": 3.0929885506629944}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:31.011043+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985411.011057, "update": 1} +{"device": "cuda", "elapsed_hours": 0.041475089144175095, "elapsed_seconds": 149.31032091903035, "env_steps_per_second": 13394.921313474251, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6175284766228288, "eta_seconds": 5823.1025158421835, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0003632648299958241, "chaser_grad_norm": 0.2142338901758194, "chaser_new_fields": 77.375, "chaser_partner_vision": 0.0, "chaser_return": 5.365001201629639, "chaser_subspace_norm": 15.691179275512695, "collisions_per_episode": 1.475000023841858, "entropy": 0.30198078602552414, "explorer_grad_norm": 0.13914613611996174, "explorer_new_fields": 70.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 54.587501525878906, "explorer_subspace_norm": 11.627245903015137, "final_distance": 4.618728160858154, "max_param_abs": 0.32930245995521545, "policy_loss": -0.0019862561977923576, "value_loss": 4.403667330741882}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:45:58.719958+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985558.7199697, "update": 500} +{"device": "cuda", "elapsed_hours": 0.09452361568750348, "elapsed_seconds": 340.2850164750125, "env_steps_per_second": 11754.851980953219, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.795948698062566, "eta_seconds": 6465.415313025238, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 7.956908544803909e-05, "chaser_grad_norm": 0.13848095387220383, "chaser_new_fields": 79.75, "chaser_partner_vision": 0.0, "chaser_return": 5.8600006103515625, "chaser_subspace_norm": 28.38010597229004, "collisions_per_episode": 1.600000023841858, "entropy": 0.18292785063385963, "explorer_grad_norm": 0.1835392415523529, "explorer_new_fields": 71.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 56.38750076293945, "explorer_subspace_norm": 15.596293449401855, "final_distance": 5.755338668823242, "max_param_abs": 0.3677484691143036, "policy_loss": -0.0005263844929217498, "value_loss": 4.1441404819488525}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:49:09.694652+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985749.6946669, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.1479966188427837, "elapsed_seconds": 532.7878278340213, "env_steps_per_second": 11261.518538049582, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8252916323943322, "eta_seconds": 6571.049876619596, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 9.463303504797516e-05, "chaser_grad_norm": 0.2207968719303608, "chaser_new_fields": 81.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.305001258850098, "chaser_subspace_norm": 35.60392379760742, "collisions_per_episode": 0.7250000238418579, "entropy": 0.18308475613594055, "explorer_grad_norm": 0.37780342996120453, "explorer_new_fields": 70.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 56.23749923706055, "explorer_subspace_norm": 17.865774154663086, "final_distance": 5.414912700653076, "max_param_abs": 0.38483133912086487, "policy_loss": -0.0008592407521845757, "value_loss": 3.9924381971359253}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:52:22.197462+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985942.1974738, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.1996872979261227, "elapsed_seconds": 718.8742725340417, "env_steps_per_second": 11128.51065291277, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7971856813351041, "eta_seconds": 6469.868452806375, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.000541657906368842, "chaser_grad_norm": 0.195209588855505, "chaser_new_fields": 79.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.830001354217529, "chaser_subspace_norm": 39.06519317626953, "collisions_per_episode": 1.350000023841858, "entropy": 0.1706830933690071, "explorer_grad_norm": 0.47785354405641556, "explorer_new_fields": 71.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 56.01250076293945, "explorer_subspace_norm": 20.26156997680664, "final_distance": 4.974127292633057, "max_param_abs": 0.4402898848056793, "policy_loss": -0.0012024157301055993, "value_loss": 4.066803336143494}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:55:28.283909+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986128.2839227, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.25255303585943895, "elapsed_seconds": 909.1909290939802, "env_steps_per_second": 10998.789891101445, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7678712510160723, "eta_seconds": 6364.336503657861, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0003182750429737702, "chaser_grad_norm": 0.11373030208051205, "chaser_new_fields": 83.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.565001010894775, "chaser_subspace_norm": 44.44955062866211, "collisions_per_episode": 0.824999988079071, "entropy": 0.16687658429145813, "explorer_grad_norm": 0.17073874920606613, "explorer_new_fields": 73.625, "explorer_partner_vision": 0.0, "explorer_return": 60.0625, "explorer_subspace_norm": 21.136234283447266, "final_distance": 4.893276691436768, "max_param_abs": 0.48829394578933716, "policy_loss": -0.0016850562803458757, "value_loss": 4.075126886367798}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:58:38.600564+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986318.6005762, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.30461443797167803, "elapsed_seconds": 1096.611976698041, "env_steps_per_second": 10942.794949342666, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.726148481839509, "eta_seconds": 6214.134534622232, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": -5.2403578119641914e-05, "chaser_grad_norm": 0.0758270900696516, "chaser_new_fields": 81.25, "chaser_partner_vision": 0.0, "chaser_return": 6.210000514984131, "chaser_subspace_norm": 52.24632263183594, "collisions_per_episode": 1.1750000715255737, "entropy": 0.142869982868433, "explorer_grad_norm": 0.41694944351911545, "explorer_new_fields": 73.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 59.6875, "explorer_subspace_norm": 23.191781997680664, "final_distance": 4.731234073638916, "max_param_abs": 0.4901277422904968, "policy_loss": -0.0007293193800719244, "value_loss": 4.039798021316528}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:01:46.021613+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986506.021627, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.35776873922027236, "elapsed_seconds": 1287.9674611929804, "env_steps_per_second": 10869.839822686587, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6866240563241413, "eta_seconds": 6071.846602766908, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 2.5298360996739433e-05, "chaser_grad_norm": 0.3993445299565792, "chaser_new_fields": 82.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.465001106262207, "chaser_subspace_norm": 55.589115142822266, "collisions_per_episode": 1.3250000476837158, "entropy": 0.1278463676571846, "explorer_grad_norm": 0.27861612290143967, "explorer_new_fields": 74.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.22500228881836, "explorer_subspace_norm": 29.22615623474121, "final_distance": 4.770143032073975, "max_param_abs": 0.5207052826881409, "policy_loss": -0.0009798301118428832, "value_loss": 3.9451226592063904}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:04:57.377099+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986697.3771112, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.41089496239001166, "elapsed_seconds": 1479.221864604042, "env_steps_per_second": 10816.4977701184, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6435798495600467, "eta_seconds": 5916.887458416168, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": -0.00013971206711469418, "chaser_grad_norm": 0.12048624269664288, "chaser_new_fields": 85.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.9500017166137695, "chaser_subspace_norm": 58.02903747558594, "collisions_per_episode": 1.75, "entropy": 0.0875592902302742, "explorer_grad_norm": 0.15481581538915634, "explorer_new_fields": 74.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 61.22500228881836, "explorer_subspace_norm": 33.89311981201172, "final_distance": 4.6245436668396, "max_param_abs": 0.5600928068161011, "policy_loss": -0.00040669443222540735, "value_loss": 3.939574360847473}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:08:08.631501+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986888.631514, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4643042968908493, "elapsed_seconds": 1671.4954688070575, "env_steps_per_second": 10768.799758007455, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.59927035595737, "eta_seconds": 5757.373281446532, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00031098692283237916, "chaser_grad_norm": 0.3256533741950989, "chaser_new_fields": 85.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.960000514984131, "chaser_subspace_norm": 60.04489517211914, "collisions_per_episode": 1.4250000715255737, "entropy": 0.09199833124876022, "explorer_grad_norm": 0.2764192707836628, "explorer_new_fields": 74.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 60.88750076293945, "explorer_subspace_norm": 37.67209243774414, "final_distance": 5.471877098083496, "max_param_abs": 0.5733164548873901, "policy_loss": -0.0009315415767474633, "value_loss": 3.910860598087311}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:11:20.905104+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987080.905115, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5159377434691609, "elapsed_seconds": 1857.375876488979, "env_steps_per_second": 10767.879702306811, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5478132304074825, "eta_seconds": 5572.127629466937, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00019433532364115935, "chaser_grad_norm": 0.17675616219639778, "chaser_new_fields": 83.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.595001220703125, "chaser_subspace_norm": 63.742034912109375, "collisions_per_episode": 1.2000000476837158, "entropy": 0.1041656993329525, "explorer_grad_norm": 0.23542096465826035, "explorer_new_fields": 74.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.337501525878906, "explorer_subspace_norm": 38.62540817260742, "final_distance": 4.564873695373535, "max_param_abs": 0.6108526587486267, "policy_loss": -0.000674923917742376, "value_loss": 3.877275824546814}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:14:26.785513+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987266.7855272, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.567737341909443, "elapsed_seconds": 2043.8544308739947, "env_steps_per_second": 10763.975979733714, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4967620832158044, "eta_seconds": 5388.343499576896, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0002979320848088135, "chaser_grad_norm": 0.3199396878480911, "chaser_new_fields": 77.625, "chaser_partner_vision": 0.0, "chaser_return": 5.465001106262207, "chaser_subspace_norm": 72.06124114990234, "collisions_per_episode": 0.925000011920929, "entropy": 0.08251870051026344, "explorer_grad_norm": 0.25879210978746414, "explorer_new_fields": 74.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 61.78750228881836, "explorer_subspace_norm": 47.33602523803711, "final_distance": 5.2332987785339355, "max_param_abs": 0.6240873336791992, "policy_loss": -0.0005941399848090612, "value_loss": 3.9337053894996643}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:17:33.264066+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987453.2640774, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6193817119269645, "elapsed_seconds": 2229.774162937072, "env_steps_per_second": 10763.421874252526, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4452239944962506, "eta_seconds": 5202.806380186502, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 7.315270792285222e-05, "chaser_grad_norm": 0.07100265473127365, "chaser_new_fields": 77.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.310000896453857, "chaser_subspace_norm": 73.39598846435547, "collisions_per_episode": 1.024999976158142, "entropy": 0.09012490883469582, "explorer_grad_norm": 0.592255488038063, "explorer_new_fields": 73.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 60.212501525878906, "explorer_subspace_norm": 48.91190719604492, "final_distance": 5.471304416656494, "max_param_abs": 0.6853839159011841, "policy_loss": -0.00046422900573839243, "value_loss": 3.929266393184662}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:20:39.183799+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987639.1838136, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6721014265913982, "elapsed_seconds": 2419.5651357290335, "env_steps_per_second": 10745.73261784333, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3959029629205961, "eta_seconds": 5025.250666514146, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006224097467039114, "chaser_grad_norm": 0.06170021556317806, "chaser_new_fields": 74.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.910000801086426, "chaser_subspace_norm": 70.08699798583984, "collisions_per_episode": 1.1750000715255737, "entropy": 0.09079068340361118, "explorer_grad_norm": 0.26018545031547546, "explorer_new_fields": 73.0, "explorer_partner_vision": 0.0, "explorer_return": 59.04999923706055, "explorer_subspace_norm": 54.48208999633789, "final_distance": 4.70918607711792, "max_param_abs": 0.7426587343215942, "policy_loss": -0.0007514058209494578, "value_loss": 3.862727642059326}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:23:48.974771+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987828.974782, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7256725682499948, "elapsed_seconds": 2612.4212456999812, "env_steps_per_second": 10718.026446189608, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3476776267499906, "eta_seconds": 4851.639456299966, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 7.923711780041809e-05, "chaser_grad_norm": 0.10907543264329433, "chaser_new_fields": 76.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.275001049041748, "chaser_subspace_norm": 86.54544830322266, "collisions_per_episode": 1.375, "entropy": 0.05557131487876177, "explorer_grad_norm": 0.24157263338565826, "explorer_new_fields": 75.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 63.17499923706055, "explorer_subspace_norm": 57.579593658447266, "final_distance": 5.452548503875732, "max_param_abs": 0.7982372641563416, "policy_loss": -0.000272475741279532, "value_loss": 3.8635100722312927}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:27:01.830882+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988021.8308973, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7782621729819544, "elapsed_seconds": 2801.743822735036, "env_steps_per_second": 10707.61707639433, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2971036216365908, "eta_seconds": 4669.573037891727, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0004368061860267197, "chaser_grad_norm": 0.14359945058822632, "chaser_new_fields": 77.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.530000686645508, "chaser_subspace_norm": 83.08934020996094, "collisions_per_episode": 1.4250000715255737, "entropy": 0.08567827567458153, "explorer_grad_norm": 0.7056173384189606, "explorer_new_fields": 76.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 64.9000015258789, "explorer_subspace_norm": 57.05202102661133, "final_distance": 4.9074883460998535, "max_param_abs": 0.8255160450935364, "policy_loss": -0.0009210895699215405, "value_loss": 3.8653560280799866}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:30:11.153457+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988211.15347, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8314077611533381, "elapsed_seconds": 2993.0679401520174, "env_steps_per_second": 10691.37107471564, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2471116417300072, "eta_seconds": 4489.601910228026, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0004440378430064229, "chaser_grad_norm": 0.11829627957195044, "chaser_new_fields": 77.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.52000093460083, "chaser_subspace_norm": 89.27839660644531, "collisions_per_episode": 1.6750000715255737, "entropy": 0.06764670088887215, "explorer_grad_norm": 0.3792828917503357, "explorer_new_fields": 74.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.45000076293945, "explorer_subspace_norm": 59.709022521972656, "final_distance": 4.851212024688721, "max_param_abs": 0.8565105199813843, "policy_loss": -0.000669933580095261, "value_loss": 3.7234076261520386}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:33:22.477577+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988402.477593, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8848842977541628, "elapsed_seconds": 3185.583471914986, "env_steps_per_second": 10673.084004783963, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1971964028438675, "eta_seconds": 4309.907050237923, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0002396710157837323, "chaser_grad_norm": 0.10101520549505949, "chaser_new_fields": 77.0, "chaser_partner_vision": 0.0, "chaser_return": 5.3600006103515625, "chaser_subspace_norm": 89.2835693359375, "collisions_per_episode": 1.0750000476837158, "entropy": 0.05736979190260172, "explorer_grad_norm": 0.2609849013388157, "explorer_new_fields": 77.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 65.42500305175781, "explorer_subspace_norm": 64.30477142333984, "final_distance": 4.804805278778076, "max_param_abs": 0.8937316536903381, "policy_loss": -0.0005577248199273299, "value_loss": 3.7547460198402405}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:36:34.993109+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988594.9931228, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9372974603680713, "elapsed_seconds": 3374.2708573250566, "env_steps_per_second": 10668.971615556344, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1455857848943094, "eta_seconds": 4124.1088256195135, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0003862881705324339, "chaser_grad_norm": 0.11785871535539627, "chaser_new_fields": 78.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.745000839233398, "chaser_subspace_norm": 87.86925506591797, "collisions_per_episode": 1.100000023841858, "entropy": 0.07302082143723965, "explorer_grad_norm": 0.13841502368450165, "explorer_new_fields": 76.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 63.70000076293945, "explorer_subspace_norm": 65.82949829101562, "final_distance": 4.995398998260498, "max_param_abs": 0.9703806638717651, "policy_loss": -0.0005471885539591881, "value_loss": 3.859369218349457}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:39:43.680493+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988783.6805072, "update": 9000} +{"device": "cuda", "elapsed_hours": 0.9899086770825023, "elapsed_seconds": 3563.6712374970084, "env_steps_per_second": 10663.16095608466, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0941095904596079, "eta_seconds": 3938.794525654588, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0003634184970375265, "chaser_grad_norm": 0.06702764425426722, "chaser_new_fields": 83.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.590001106262207, "chaser_subspace_norm": 98.16253662109375, "collisions_per_episode": 0.9750000238418579, "entropy": 0.07850857824087143, "explorer_grad_norm": 0.25295310094952583, "explorer_new_fields": 74.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.1875, "explorer_subspace_norm": 66.24516296386719, "final_distance": 4.304412841796875, "max_param_abs": 0.9871649146080017, "policy_loss": -0.0011359985345045231, "value_loss": 3.8722885251045227}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:42:53.080873+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988973.0808852, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0430193441636058, "elapsed_seconds": 3754.869638988981, "env_steps_per_second": 10652.833212811674, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0430193441636058, "eta_seconds": 3754.869638988981, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0003298464110783588, "chaser_grad_norm": 0.08921568095684052, "chaser_new_fields": 80.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.060000896453857, "chaser_subspace_norm": 93.15203094482422, "collisions_per_episode": 1.0500000715255737, "entropy": 0.06284694839268923, "explorer_grad_norm": 0.6557697653770447, "explorer_new_fields": 76.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 64.0, "explorer_subspace_norm": 71.28758239746094, "final_distance": 5.206691265106201, "max_param_abs": 1.0260392427444458, "policy_loss": -0.0005279974394944986, "value_loss": 3.7852091193199158}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:46:04.279276+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989164.2792902, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.0948423628338948, "elapsed_seconds": 3941.4325062020216, "env_steps_per_second": 10656.024157184249, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9905716616116192, "eta_seconds": 3566.057981801829, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 3.462782712038669e-05, "chaser_grad_norm": 0.07610498368740082, "chaser_new_fields": 77.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.385000705718994, "chaser_subspace_norm": 84.84783172607422, "collisions_per_episode": 0.9000000357627869, "entropy": 0.05232377257198095, "explorer_grad_norm": 0.1455609779804945, "explorer_new_fields": 75.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 63.17499923706055, "explorer_subspace_norm": 83.44246673583984, "final_distance": 5.265476226806641, "max_param_abs": 1.0845651626586914, "policy_loss": -0.00020808578724906823, "value_loss": 3.767380177974701}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:49:10.842142+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989350.842152, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1489696305983428, "elapsed_seconds": 4136.290670154034, "env_steps_per_second": 10637.550285691468, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9400660613986441, "eta_seconds": 3384.237821035119, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 7.519029446321959e-05, "chaser_grad_norm": 0.0776878697797656, "chaser_new_fields": 79.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.7900004386901855, "chaser_subspace_norm": 100.6033706665039, "collisions_per_episode": 1.1750000715255737, "entropy": 0.0600092438980937, "explorer_grad_norm": 0.19327158108353615, "explorer_new_fields": 74.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.22500228881836, "explorer_subspace_norm": 85.98049926757812, "final_distance": 5.141221046447754, "max_param_abs": 1.1063262224197388, "policy_loss": -0.00039013837714452215, "value_loss": 3.7150986790657043}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:52:25.700306+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989545.7003198, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2020321397597178, "elapsed_seconds": 4327.315703134984, "env_steps_per_second": 10630.146528637757, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8884585380832696, "eta_seconds": 3198.4507370997703, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0003250303412194108, "chaser_grad_norm": 0.04050362715497613, "chaser_new_fields": 80.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.075000762939453, "chaser_subspace_norm": 93.22357177734375, "collisions_per_episode": 1.3000000715255737, "entropy": 0.09896882250905037, "explorer_grad_norm": 0.27636435627937317, "explorer_new_fields": 74.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 61.22500228881836, "explorer_subspace_norm": 72.65573120117188, "final_distance": 5.30914831161499, "max_param_abs": 1.1300534009933472, "policy_loss": -0.0006692521995272926, "value_loss": 3.807954490184784}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:55:36.725338+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989736.725349, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2538694564463932, "elapsed_seconds": 4513.930043207016, "env_steps_per_second": 10633.74920314392, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8359129709642621, "eta_seconds": 3009.2866954713436, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0001460757644520319, "chaser_grad_norm": 0.09414643235504627, "chaser_new_fields": 76.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.325000286102295, "chaser_subspace_norm": 87.30382537841797, "collisions_per_episode": 1.2000000476837158, "entropy": 0.06684242933988571, "explorer_grad_norm": 0.36964280903339386, "explorer_new_fields": 76.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 64.82500457763672, "explorer_subspace_norm": 87.4162368774414, "final_distance": 4.528558254241943, "max_param_abs": 1.1710652112960815, "policy_loss": -0.0006748977715602678, "value_loss": 3.8218905925750732}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:58:43.339680+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989923.339697, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.306414859188048, "elapsed_seconds": 4703.093493076973, "env_steps_per_second": 10631.300456518838, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7838489155128288, "eta_seconds": 2821.8560958461835, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0001932052191510325, "chaser_grad_norm": 0.06965518184006214, "chaser_new_fields": 76.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.135000705718994, "chaser_subspace_norm": 91.65786743164062, "collisions_per_episode": 1.3250000476837158, "entropy": 0.06623650528490543, "explorer_grad_norm": 0.5132623240351677, "explorer_new_fields": 76.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 63.4375, "explorer_subspace_norm": 86.87879943847656, "final_distance": 5.124466896057129, "max_param_abs": 1.2168129682540894, "policy_loss": -0.0004816767695601243, "value_loss": 3.904612720012665}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:01:52.503129+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990112.503141, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.359832654665289, "elapsed_seconds": 4895.39755679504, "env_steps_per_second": 10622.222076289101, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7322175832813094, "eta_seconds": 2635.9832998127135, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -0.0002154850242672146, "chaser_grad_norm": 0.10619407892227173, "chaser_new_fields": 76.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.3000006675720215, "chaser_subspace_norm": 94.92156219482422, "collisions_per_episode": 0.824999988079071, "entropy": 0.08630450069904327, "explorer_grad_norm": 0.27689390257000923, "explorer_new_fields": 74.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.337501525878906, "explorer_subspace_norm": 89.77205657958984, "final_distance": 4.834729194641113, "max_param_abs": 1.2447625398635864, "policy_loss": -0.0008252897658382885, "value_loss": 3.7048959732055664}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:05:04.807193+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990304.807207, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4118516931288954, "elapsed_seconds": 5082.666095264023, "env_steps_per_second": 10624.34537069367, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6797804448398385, "eta_seconds": 2447.2096014234185, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -0.0002211178556340448, "chaser_grad_norm": 0.09127848222851753, "chaser_new_fields": 74.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.780000686645508, "chaser_subspace_norm": 105.30447387695312, "collisions_per_episode": 0.699999988079071, "entropy": 0.05811640154570341, "explorer_grad_norm": 0.2383578047156334, "explorer_new_fields": 76.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 64.7125015258789, "explorer_subspace_norm": 87.39847564697266, "final_distance": 4.836423397064209, "max_param_abs": 1.259665608406067, "policy_loss": -0.000673688062700073, "value_loss": 3.7406721711158752}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:08:12.075731+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990492.0757418, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4654038873622388, "elapsed_seconds": 5275.45399450406, "env_steps_per_second": 10615.200143597216, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6280302374409594, "eta_seconds": 2260.908854787454, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0004374934122642893, "chaser_grad_norm": 0.18387733399868011, "chaser_new_fields": 71.75, "chaser_partner_vision": 0.0, "chaser_return": 4.3050007820129395, "chaser_subspace_norm": 87.46336364746094, "collisions_per_episode": 0.8500000238418579, "entropy": 0.06856168061494827, "explorer_grad_norm": 0.4256523996591568, "explorer_new_fields": 74.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 60.47500228881836, "explorer_subspace_norm": 91.64143371582031, "final_distance": 4.771363735198975, "max_param_abs": 1.274393916130066, "policy_loss": -0.0007815306585978732, "value_loss": 3.6483676433563232}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:11:24.863630+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990684.8636446, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5174561206619384, "elapsed_seconds": 5462.842034382978, "env_steps_per_second": 10617.184175370547, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5755868043890112, "eta_seconds": 2072.1124958004402, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0010260396141429373, "chaser_grad_norm": 0.1073277797549963, "chaser_new_fields": 72.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.470000743865967, "chaser_subspace_norm": 92.8544921875, "collisions_per_episode": 0.949999988079071, "entropy": 0.10074484907090664, "explorer_grad_norm": 0.9042353183031082, "explorer_new_fields": 74.625, "explorer_partner_vision": 0.0, "explorer_return": 61.525001525878906, "explorer_subspace_norm": 97.54423522949219, "final_distance": 4.749428749084473, "max_param_abs": 1.2983474731445312, "policy_loss": -1.331270125159989e-05, "value_loss": 3.6823132038116455}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:14:32.251670+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990872.2516806, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5706263935241602, "elapsed_seconds": 5654.255016686977, "env_steps_per_second": 10611.477519660242, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.52354213117472, "eta_seconds": 1884.7516722289922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005047082217448096, "chaser_grad_norm": 0.06792480684816837, "chaser_new_fields": 73.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 4.6700005531311035, "chaser_subspace_norm": 93.45286560058594, "collisions_per_episode": 1.4250000715255737, "entropy": 0.0717815849930048, "explorer_grad_norm": 0.44703781604766846, "explorer_new_fields": 75.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.53750228881836, "explorer_subspace_norm": 98.89013671875, "final_distance": 4.489287376403809, "max_param_abs": 1.2968024015426636, "policy_loss": -0.0009549594561106822, "value_loss": 3.677554965019226}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:17:43.664652+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991063.6646657, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6228326400964095, "elapsed_seconds": 5842.197504347074, "env_steps_per_second": 10612.445052374029, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47114496002798983, "eta_seconds": 1696.1218561007634, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00027427483835661126, "chaser_grad_norm": 0.038474843837320805, "chaser_new_fields": 75.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.960000514984131, "chaser_subspace_norm": 100.0431137084961, "collisions_per_episode": 1.25, "entropy": 0.06876423768699169, "explorer_grad_norm": 0.45563749969005585, "explorer_new_fields": 73.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.57500076293945, "explorer_subspace_norm": 100.91221618652344, "final_distance": 4.574155330657959, "max_param_abs": 1.3337393999099731, "policy_loss": -0.0007041651313772768, "value_loss": 3.8305552005767822}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:20:51.607139+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991251.6071503, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6750371870913983, "elapsed_seconds": 6030.133873529034, "env_steps_per_second": 10613.363043388801, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4187592967728495, "eta_seconds": 1507.5334683822582, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0004321018841346863, "chaser_grad_norm": 0.049426293931901455, "chaser_new_fields": 74.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.905000686645508, "chaser_subspace_norm": 102.40481567382812, "collisions_per_episode": 1.100000023841858, "entropy": 0.0732938889414072, "explorer_grad_norm": 0.8978658765554428, "explorer_new_fields": 74.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.5625, "explorer_subspace_norm": 114.98880767822266, "final_distance": 3.9745118618011475, "max_param_abs": 1.335937261581421, "policy_loss": -0.00044449617178088374, "value_loss": 3.6162365674972534}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:23:59.543510+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991439.543524, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.728043479719458, "elapsed_seconds": 6220.956526990049, "env_steps_per_second": 10609.300951333522, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.36655467751624865, "eta_seconds": 1319.596839058495, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0004602538025242908, "chaser_grad_norm": 0.0318006738089025, "chaser_new_fields": 74.875, "chaser_partner_vision": 0.0, "chaser_return": 4.880000591278076, "chaser_subspace_norm": 93.14822387695312, "collisions_per_episode": 1.4250000715255737, "entropy": 0.06661080196499825, "explorer_grad_norm": 0.7576425969600677, "explorer_new_fields": 73.375, "explorer_partner_vision": 0.0, "explorer_return": 59.3125, "explorer_subspace_norm": 109.37227630615234, "final_distance": 4.762907028198242, "max_param_abs": 1.3563886880874634, "policy_loss": -0.0004989569962490226, "value_loss": 3.796582877635956}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:27:10.366164+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991630.3661783, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7815882814236101, "elapsed_seconds": 6413.717813124997, "env_steps_per_second": 10602.274995766913, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31439793201593125, "eta_seconds": 1131.8325552573524, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0003217695078456262, "chaser_grad_norm": 0.13442767038941383, "chaser_new_fields": 77.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.410000801086426, "chaser_subspace_norm": 105.91724395751953, "collisions_per_episode": 1.625, "entropy": 0.0788711104542017, "explorer_grad_norm": 0.38502760231494904, "explorer_new_fields": 72.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 58.67499923706055, "explorer_subspace_norm": 99.43160247802734, "final_distance": 5.037087917327881, "max_param_abs": 1.3281008005142212, "policy_loss": -0.0007574277424717835, "value_loss": 3.580711305141449}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:30:23.127451+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991823.127467, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8352877387569364, "elapsed_seconds": 6607.035859524971, "env_steps_per_second": 10594.766168717726, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26218396267956234, "eta_seconds": 943.8622656464244, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0007609897645764613, "chaser_grad_norm": 0.11335813254117966, "chaser_new_fields": 75.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 4.955000877380371, "chaser_subspace_norm": 104.22798919677734, "collisions_per_episode": 1.125, "entropy": 0.06128899473696947, "explorer_grad_norm": 0.44147511944174767, "explorer_new_fields": 75.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 62.98749923706055, "explorer_subspace_norm": 105.55374145507812, "final_distance": 5.476316928863525, "max_param_abs": 1.3470975160598755, "policy_loss": -0.000845771472784751, "value_loss": 3.771251142024994}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:33:36.445496+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992016.445509, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.8888409559169668, "elapsed_seconds": 6799.82744130108, "env_steps_per_second": 10588.503990951791, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.20987121732410746, "eta_seconds": 755.5363823667868, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0004439944464451895, "chaser_grad_norm": 0.09758789651095867, "chaser_new_fields": 76.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.2400007247924805, "chaser_subspace_norm": 100.56327819824219, "collisions_per_episode": 1.225000023841858, "entropy": 0.07399779930710793, "explorer_grad_norm": 0.3509122431278229, "explorer_new_fields": 74.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.41250228881836, "explorer_subspace_norm": 118.36441040039062, "final_distance": 4.870284080505371, "max_param_abs": 1.4301035404205322, "policy_loss": -0.0010395317695799378, "value_loss": 3.697541356086731}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:36:49.237078+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992209.2370934, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9412984795449948, "elapsed_seconds": 6988.674526361981, "env_steps_per_second": 10588.5600654122, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15740257942256713, "eta_seconds": 566.6492859212417, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00041806527268874183, "chaser_grad_norm": 0.06867756694555283, "chaser_new_fields": 79.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.7350006103515625, "chaser_subspace_norm": 95.10716247558594, "collisions_per_episode": 1.524999976158142, "entropy": 0.06634167023003101, "explorer_grad_norm": 0.3984520249068737, "explorer_new_fields": 74.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.07500076293945, "explorer_subspace_norm": 112.43339538574219, "final_distance": 5.087775707244873, "max_param_abs": 1.606360912322998, "policy_loss": -0.0007691143096892894, "value_loss": 3.678083837032318}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:39:58.084161+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992398.084172, "update": 18500} +{"device": "cuda", "elapsed_hours": 1.9943824925714095, "elapsed_seconds": 7179.776973257074, "env_steps_per_second": 10585.28701978927, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10496749960902155, "eta_seconds": 377.8829985924776, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00028678630876688427, "chaser_grad_norm": 0.10852688178420067, "chaser_new_fields": 76.875, "chaser_partner_vision": 0.0, "chaser_return": 5.315001010894775, "chaser_subspace_norm": 94.96185302734375, "collisions_per_episode": 1.3250000476837158, "entropy": 0.07276676408946514, "explorer_grad_norm": 0.6109379157423973, "explorer_new_fields": 75.5, "explorer_partner_vision": 0.0, "explorer_return": 62.61249923706055, "explorer_subspace_norm": 117.20232391357422, "final_distance": 4.531687259674072, "max_param_abs": 1.598679542541504, "policy_loss": -0.00040980818419456, "value_loss": 3.7027546763420105}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:43:09.186609+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992589.186624, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.04765860068919, "elapsed_seconds": 7371.570962481084, "env_steps_per_second": 10581.190956038383, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0525040666843382, "eta_seconds": 189.01464006361752, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005783102995236961, "chaser_grad_norm": 0.13811812549829483, "chaser_new_fields": 77.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.460000991821289, "chaser_subspace_norm": 99.20040893554688, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0773265864700079, "explorer_grad_norm": 0.30666641891002655, "explorer_new_fields": 75.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 63.32500076293945, "explorer_subspace_norm": 108.5719985961914, "final_distance": 4.436137676239014, "max_param_abs": 1.6025590896606445, "policy_loss": -0.0009690576277208152, "value_loss": 3.7210079431533813}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:46:20.980598+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992780.9806087, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.100639237591402, "elapsed_seconds": 7562.301255329046, "env_steps_per_second": 10578.790410343034, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:49:31.710891+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992971.7109015, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1006427376291765, "elapsed_seconds": 7562.313855465036, "env_steps_per_second": 10578.772784230137, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:49:31.723491+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992971.7235026, "update": 20000} diff --git a/non_social/seed_0005/attempt_01/status.json b/non_social/seed_0005/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..b292ee4043e60eb45d9dfd838019ea81b7d54789 --- /dev/null +++ b/non_social/seed_0005/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1006427376291765, + "elapsed_seconds": 7562.313855465036, + "env_steps_per_second": 10578.772784230137, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.00022429173202174373, + "chaser_grad_norm": 0.04249562695622444, + "chaser_new_fields": 80.4749984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.015000820159912, + "chaser_subspace_norm": 93.73216247558594, + "collisions_per_episode": 1.024999976158142, + "entropy": 0.05529557354748249, + "explorer_grad_norm": 0.39152009785175323, + "explorer_new_fields": 73.2249984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 59.42499923706055, + "explorer_subspace_norm": 117.07512664794922, + "final_distance": 4.730400562286377, + "max_param_abs": 1.606895923614502, + "policy_loss": -0.0002411644566997495, + "value_loss": 3.523293375968933 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:49:31.723491+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782992971.7235026, + "update": 20000 +} diff --git a/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..47a492c3b183b2e668fb787160defaef39071ae1 --- /dev/null +++ b/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01b8815629b4e75ed7baddd26bfade4a209f1c7c07e34b4bc26af47afcf2b7ed +size 11460384 diff --git a/non_social/seed_0006/attempt_01/config.json b/non_social/seed_0006/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7e990c9c27e8a6bb46d9071cee6f124560363ac3 --- /dev/null +++ b/non_social/seed_0006/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", + "seed": 6, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0006/attempt_01/metrics.jsonl b/non_social/seed_0006/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..cd0b15b159d2e33436af6d3b69c1e80cb81b1c17 --- /dev/null +++ b/non_social/seed_0006/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0004029932147305873, "elapsed_seconds": 1.4507755730301142, "env_steps_per_second": 2757.1459530749703, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 8.059461301397015, "eta_seconds": 29014.06068502925, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 4.250772384722934e-06, "chaser_grad_norm": 0.4149998426437378, "chaser_new_fields": 34.400001525878906, "chaser_partner_vision": 0.0, "chaser_return": -3.200000047683716, "chaser_subspace_norm": 0.0025991403963416815, "collisions_per_episode": 2.3500001430511475, "entropy": 2.772491693496704, "explorer_grad_norm": 0.11106047220528126, "explorer_new_fields": 33.45000076293945, "explorer_partner_vision": 0.0, "explorer_return": -0.42500001192092896, "explorer_subspace_norm": 0.0026027511339634657, "final_distance": 4.608747959136963, "max_param_abs": 0.1542823314666748, "policy_loss": -0.0005767088214643223, "value_loss": 3.029911696910858}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:50.905127+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985430.905139, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05063779645916333, "elapsed_seconds": 182.296067252988, "env_steps_per_second": 10971.163723594911, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.97487406190737, "eta_seconds": 7109.546622866532, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00018145570814898093, "chaser_grad_norm": 0.15711118653416634, "chaser_new_fields": 75.625, "chaser_partner_vision": 0.0, "chaser_return": 5.090001106262207, "chaser_subspace_norm": 14.87479305267334, "collisions_per_episode": 0.75, "entropy": 0.3481130003929138, "explorer_grad_norm": 0.23191631212830544, "explorer_new_fields": 62.47500228881836, "explorer_partner_vision": 0.0, "explorer_return": 43.26250076293945, "explorer_subspace_norm": 13.244866371154785, "final_distance": 5.966777324676514, "max_param_abs": 0.31924331188201904, "policy_loss": -0.0021471149744449747, "value_loss": 4.699725985527039}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:46:51.750418+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985611.75043, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10273150985222956, "elapsed_seconds": 369.83343546802644, "env_steps_per_second": 10815.679753070395, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9518986871923618, "eta_seconds": 7026.835273892502, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00019757886701632366, "chaser_grad_norm": 0.23910705372691154, "chaser_new_fields": 79.375, "chaser_partner_vision": 0.0, "chaser_return": 5.830000877380371, "chaser_subspace_norm": 26.12096405029297, "collisions_per_episode": 0.949999988079071, "entropy": 0.26529232412576675, "explorer_grad_norm": 0.5567674934864044, "explorer_new_fields": 60.900001525878906, "explorer_partner_vision": 0.0, "explorer_return": 40.900001525878906, "explorer_subspace_norm": 14.061508178710938, "final_distance": 5.95941162109375, "max_param_abs": 0.34445086121559143, "policy_loss": -0.0014283773567242974, "value_loss": 4.23380172252655}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:49:59.287786+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985799.2878, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15487551840917957, "elapsed_seconds": 557.5518662730465, "env_steps_per_second": 10761.330672439462, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9101313937132145, "eta_seconds": 6876.4730173675725, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0005860907181768349, "chaser_grad_norm": 0.09903131052851677, "chaser_new_fields": 80.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.0500006675720215, "chaser_subspace_norm": 33.27919006347656, "collisions_per_episode": 0.9000000357627869, "entropy": 0.2126907855272293, "explorer_grad_norm": 0.5906994789838791, "explorer_new_fields": 59.47500228881836, "explorer_partner_vision": 0.0, "explorer_return": 38.837501525878906, "explorer_subspace_norm": 15.013877868652344, "final_distance": 5.933682441711426, "max_param_abs": 0.3797305226325989, "policy_loss": -0.0012547551710149207, "value_loss": 4.258698225021362}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:07.006216+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985987.00623, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.20882671108889755, "elapsed_seconds": 751.7761599200312, "env_steps_per_second": 10641.46540753698, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8794403998000777, "eta_seconds": 6765.98543928028, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00018122987076776842, "chaser_grad_norm": 0.20974123291671276, "chaser_new_fields": 79.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.835000514984131, "chaser_subspace_norm": 40.17341613769531, "collisions_per_episode": 0.6000000238418579, "entropy": 0.19857148453593254, "explorer_grad_norm": 0.8207572065293789, "explorer_new_fields": 59.72500228881836, "explorer_partner_vision": 0.0, "explorer_return": 39.36249923706055, "explorer_subspace_norm": 14.959667205810547, "final_distance": 5.517897129058838, "max_param_abs": 0.3975472152233124, "policy_loss": -0.0005522465289935496, "value_loss": 4.0377117395401}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:21.230510+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986181.230523, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26160411919444515, "elapsed_seconds": 941.7748291000025, "env_steps_per_second": 10618.249385106627, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.831228834361116, "eta_seconds": 6592.423803700018, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 8.404761288915985e-05, "chaser_grad_norm": 0.13433636352419853, "chaser_new_fields": 84.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.87000036239624, "chaser_subspace_norm": 45.45563888549805, "collisions_per_episode": 0.5250000357627869, "entropy": 0.15066828206181526, "explorer_grad_norm": 0.640210434794426, "explorer_new_fields": 61.17499923706055, "explorer_partner_vision": 0.0, "explorer_return": 41.57500076293945, "explorer_subspace_norm": 16.097702026367188, "final_distance": 5.537438869476318, "max_param_abs": 0.42714041471481323, "policy_loss": -0.0004748223744712732, "value_loss": 3.7812684774398804}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:31.229180+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986371.229195, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.314048101703908, "elapsed_seconds": 1130.5731661340687, "env_steps_per_second": 10614.085279445755, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7796059096554788, "eta_seconds": 6406.581274759724, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": -7.054970865084421e-05, "chaser_grad_norm": 0.0702746007591486, "chaser_new_fields": 82.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.365001201629639, "chaser_subspace_norm": 42.15382766723633, "collisions_per_episode": 0.625, "entropy": 0.12986303120851517, "explorer_grad_norm": 0.38760821521282196, "explorer_new_fields": 62.650001525878906, "explorer_partner_vision": 0.0, "explorer_return": 43.82500076293945, "explorer_subspace_norm": 17.92761993408203, "final_distance": 5.615963935852051, "max_param_abs": 0.4817543029785156, "policy_loss": -0.0006814771958119081, "value_loss": 4.087894558906555}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:02:40.027516+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986560.027531, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.36699247529888657, "elapsed_seconds": 1321.1729110759916, "env_steps_per_second": 10596.644756058538, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7301073835518939, "eta_seconds": 6228.386580786818, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00014498687113895237, "chaser_grad_norm": 0.07451495341956615, "chaser_new_fields": 84.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.870001316070557, "chaser_subspace_norm": 42.957454681396484, "collisions_per_episode": 0.6500000357627869, "entropy": 0.12198884226381779, "explorer_grad_norm": 0.7957862541079521, "explorer_new_fields": 59.375, "explorer_partner_vision": 0.0, "explorer_return": 38.95000076293945, "explorer_subspace_norm": 17.464141845703125, "final_distance": 5.965550899505615, "max_param_abs": 0.5363050103187561, "policy_loss": -0.0004670726146969173, "value_loss": 3.9271827340126038}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:05:50.627260+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986750.627271, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.41867698623805255, "elapsed_seconds": 1507.2371504569892, "env_steps_per_second": 10615.449596069771, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6747079449522102, "eta_seconds": 6028.948601827957, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": -1.483792714440657e-05, "chaser_grad_norm": 0.22081849724054337, "chaser_new_fields": 84.25, "chaser_partner_vision": 0.0, "chaser_return": 6.800001621246338, "chaser_subspace_norm": 50.51359939575195, "collisions_per_episode": 0.875, "entropy": 0.11821207776665688, "explorer_grad_norm": 0.6679219529032707, "explorer_new_fields": 62.70000076293945, "explorer_partner_vision": 0.0, "explorer_return": 43.75, "explorer_subspace_norm": 18.930234909057617, "final_distance": 5.365249156951904, "max_param_abs": 0.5532652735710144, "policy_loss": -0.0011956564483703858, "value_loss": 3.941063165664673}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:08:56.691500+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986936.6915152, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4705707972013948, "elapsed_seconds": 1694.0548699250212, "env_steps_per_second": 10625.393734027446, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6208549681381377, "eta_seconds": 5835.077885297295, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0002687456862965898, "chaser_grad_norm": 0.08690165169537067, "chaser_new_fields": 83.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.630001068115234, "chaser_subspace_norm": 49.43014144897461, "collisions_per_episode": 0.625, "entropy": 0.14277959987521172, "explorer_grad_norm": 0.452303983271122, "explorer_new_fields": 61.17499923706055, "explorer_partner_vision": 0.0, "explorer_return": 41.61249923706055, "explorer_subspace_norm": 18.562118530273438, "final_distance": 5.819535255432129, "max_param_abs": 0.5873720645904541, "policy_loss": -0.0011473000445363368, "value_loss": 3.978592097759247}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:03.509218+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987123.5092301, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5223369852747419, "elapsed_seconds": 1880.4131469890708, "env_steps_per_second": 10635.96052390089, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5670109558242258, "eta_seconds": 5641.239440967212, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00010295512317276545, "chaser_grad_norm": 0.09772861190140247, "chaser_new_fields": 82.75, "chaser_partner_vision": 0.0, "chaser_return": 6.465001106262207, "chaser_subspace_norm": 52.70188522338867, "collisions_per_episode": 1.4250000715255737, "entropy": 0.09814230352640152, "explorer_grad_norm": 1.937050461769104, "explorer_new_fields": 59.375, "explorer_partner_vision": 0.0, "explorer_return": 38.35000228881836, "explorer_subspace_norm": 20.620901107788086, "final_distance": 4.955056667327881, "max_param_abs": 0.5866839289665222, "policy_loss": -0.0009586477167999874, "value_loss": 3.917921245098114}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:09.867496+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987309.86751, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5753582458180608, "elapsed_seconds": 2071.289684945019, "env_steps_per_second": 10621.401805795202, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.516853557156706, "eta_seconds": 5460.672805764141, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": -1.6789954626894588e-05, "chaser_grad_norm": 0.06099529471248388, "chaser_new_fields": 80.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.120000839233398, "chaser_subspace_norm": 52.8476448059082, "collisions_per_episode": 1.100000023841858, "entropy": 0.08361699804663658, "explorer_grad_norm": 1.3672637343406677, "explorer_new_fields": 62.07500076293945, "explorer_partner_vision": 0.0, "explorer_return": 42.8125, "explorer_subspace_norm": 22.787059783935547, "final_distance": 4.762673377990723, "max_param_abs": 0.6090776324272156, "policy_loss": -0.000572077293526263, "value_loss": 3.6622779965400696}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:18:20.744036+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987500.7440493, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6282044861275224, "elapsed_seconds": 2261.5361500590807, "env_steps_per_second": 10612.255744562395, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4658104676308856, "eta_seconds": 5276.917683471188, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006370787510379394, "chaser_grad_norm": 0.05119013227522373, "chaser_new_fields": 83.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.645000457763672, "chaser_subspace_norm": 55.78033447265625, "collisions_per_episode": 1.225000023841858, "entropy": 0.1108908299356699, "explorer_grad_norm": 0.5062844753265381, "explorer_new_fields": 61.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 41.76250076293945, "explorer_subspace_norm": 25.04187774658203, "final_distance": 5.4328203201293945, "max_param_abs": 0.616727888584137, "policy_loss": -0.000730912977363829, "value_loss": 3.9319620728492737}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:21:30.990500+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987690.990515, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6797884290000203, "elapsed_seconds": 2447.2383444000734, "env_steps_per_second": 10624.22058705269, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4118682756154268, "eta_seconds": 5082.725792215537, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003191942762594824, "chaser_grad_norm": 0.1923944279551506, "chaser_new_fields": 83.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.620001316070557, "chaser_subspace_norm": 55.56820297241211, "collisions_per_episode": 0.7750000357627869, "entropy": 0.08682228066027164, "explorer_grad_norm": 0.5557020008563995, "explorer_new_fields": 63.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 45.025001525878906, "explorer_subspace_norm": 27.01348304748535, "final_distance": 5.0750651359558105, "max_param_abs": 0.6283434629440308, "policy_loss": -0.00046653214993419567, "value_loss": 3.8291852474212646}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:24:36.692693+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987876.692705, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7323891549572322, "elapsed_seconds": 2636.600957846036, "env_steps_per_second": 10619.73368274679, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.360151287777717, "eta_seconds": 4896.544635999781, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00027965294301657284, "chaser_grad_norm": 0.15538210794329643, "chaser_new_fields": 82.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.355000972747803, "chaser_subspace_norm": 55.79527282714844, "collisions_per_episode": 1.274999976158142, "entropy": 0.10262735933065414, "explorer_grad_norm": 0.30196429044008255, "explorer_new_fields": 63.650001525878906, "explorer_partner_vision": 0.0, "explorer_return": 45.10000228881836, "explorer_subspace_norm": 24.745519638061523, "final_distance": 5.443922996520996, "max_param_abs": 0.709766149520874, "policy_loss": -0.0009439783187161765, "value_loss": 3.814740002155304}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:27:46.055309+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988066.0553238, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7852050600105819, "elapsed_seconds": 2826.738216038095, "env_steps_per_second": 10612.938909513687, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3086751000176364, "eta_seconds": 4711.230360063491, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -6.500925358728094e-05, "chaser_grad_norm": 0.10574692580848932, "chaser_new_fields": 78.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.680001258850098, "chaser_subspace_norm": 62.65061569213867, "collisions_per_episode": 1.0500000715255737, "entropy": 0.07318281382322311, "explorer_grad_norm": 0.4011020362377167, "explorer_new_fields": 64.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 46.525001525878906, "explorer_subspace_norm": 29.17282485961914, "final_distance": 5.278419017791748, "max_param_abs": 0.7741085886955261, "policy_loss": -0.0007383784516719061, "value_loss": 3.8742582201957703}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:30:56.192565+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988256.192576, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8385489163730785, "elapsed_seconds": 3018.7760989430826, "env_steps_per_second": 10600.3224323936, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2578233745596177, "eta_seconds": 4528.164148414624, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.000746529453969913, "chaser_grad_norm": 0.11818954348564148, "chaser_new_fields": 78.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.665000915527344, "chaser_subspace_norm": 63.52378845214844, "collisions_per_episode": 0.699999988079071, "entropy": 0.07904841750860214, "explorer_grad_norm": 0.48911333456635475, "explorer_new_fields": 64.375, "explorer_partner_vision": 0.0, "explorer_return": 46.1875, "explorer_subspace_norm": 31.520326614379883, "final_distance": 4.727902889251709, "max_param_abs": 0.8666531443595886, "policy_loss": -0.0011373912433612077, "value_loss": 3.9685007333755493}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:34:08.230448+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988448.2304626, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8915584649647482, "elapsed_seconds": 3209.6104738730937, "env_steps_per_second": 10593.185770288068, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2062261584817182, "eta_seconds": 4342.414170534185, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 7.471377692220074e-05, "chaser_grad_norm": 0.12246733717620373, "chaser_new_fields": 83.0, "chaser_partner_vision": 0.0, "chaser_return": 6.500000953674316, "chaser_subspace_norm": 65.81596374511719, "collisions_per_episode": 0.9750000238418579, "entropy": 0.052849480882287025, "explorer_grad_norm": 0.28052495047450066, "explorer_new_fields": 68.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 53.04999923706055, "explorer_subspace_norm": 35.60272216796875, "final_distance": 4.713712215423584, "max_param_abs": 0.8675143122673035, "policy_loss": -0.0006151877040387888, "value_loss": 3.681655466556549}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:37:19.064823+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988639.0648344, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9449944877955648, "elapsed_seconds": 3401.9801560640335, "env_steps_per_second": 10582.072307455985, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1549932628612458, "eta_seconds": 4157.975746300485, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0003037717343690183, "chaser_grad_norm": 0.36215347051620483, "chaser_new_fields": 78.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.610001087188721, "chaser_subspace_norm": 57.98796844482422, "collisions_per_episode": 1.100000023841858, "entropy": 0.042170217260718346, "explorer_grad_norm": 0.4060757979750633, "explorer_new_fields": 70.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 55.11249923706055, "explorer_subspace_norm": 37.98765182495117, "final_distance": 4.9229021072387695, "max_param_abs": 0.8675143122673035, "policy_loss": -0.000815018462482886, "value_loss": 3.8162490129470825}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:40:31.434506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988831.4345205, "update": 9000} +{"device": "cuda", "elapsed_hours": 0.9983434119489134, "elapsed_seconds": 3594.0362830160884, "env_steps_per_second": 10573.07077827013, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1034321921540624, "eta_seconds": 3972.3558917546243, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 9.786168239181436e-05, "chaser_grad_norm": 0.13110273703932762, "chaser_new_fields": 80.625, "chaser_partner_vision": 0.0, "chaser_return": 6.090001106262207, "chaser_subspace_norm": 61.80338668823242, "collisions_per_episode": 0.699999988079071, "entropy": 0.06535376980900764, "explorer_grad_norm": 0.5733939483761787, "explorer_new_fields": 66.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 48.962501525878906, "explorer_subspace_norm": 37.62089157104492, "final_distance": 5.170069694519043, "max_param_abs": 0.8675143122673035, "policy_loss": -0.0005387380408912179, "value_loss": 3.700093626976013}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:43:43.490632+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989023.4906425, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0519645368227954, "elapsed_seconds": 3787.072332562064, "env_steps_per_second": 10562.248747157899, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0519645368227954, "eta_seconds": 3787.072332562064, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00018475628437930602, "chaser_grad_norm": 0.08554710261523724, "chaser_new_fields": 79.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.805001258850098, "chaser_subspace_norm": 68.06903839111328, "collisions_per_episode": 0.699999988079071, "entropy": 0.04232914838939905, "explorer_grad_norm": 1.0137977749109268, "explorer_new_fields": 69.625, "explorer_partner_vision": 0.0, "explorer_return": 54.25, "explorer_subspace_norm": 38.58361053466797, "final_distance": 4.837390422821045, "max_param_abs": 0.8675157427787781, "policy_loss": 5.8877678836388725e-05, "value_loss": 3.7195927500724792}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:46:56.526683+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989216.5266984, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.104521916516662, "elapsed_seconds": 3976.278899459983, "env_steps_per_second": 10562.639357542048, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9993293530388847, "eta_seconds": 3597.585670939985, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00036096852823697123, "chaser_grad_norm": 0.07153958827257156, "chaser_new_fields": 81.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.245000839233398, "chaser_subspace_norm": 65.13768005371094, "collisions_per_episode": 1.0750000476837158, "entropy": 0.0525373350828886, "explorer_grad_norm": 0.6234023571014404, "explorer_new_fields": 68.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 52.86249923706055, "explorer_subspace_norm": 38.17926025390625, "final_distance": 4.621270656585693, "max_param_abs": 0.8980355262756348, "policy_loss": -0.0009414202584889786, "value_loss": 3.4815558195114136}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:05.733248+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989405.7332592, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1575932193983398, "elapsed_seconds": 4167.335589834023, "env_steps_per_second": 10558.3049532501, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9471217249622781, "eta_seconds": 3409.638209864201, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00011893494170067687, "chaser_grad_norm": 0.09257739596068859, "chaser_new_fields": 80.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.150001049041748, "chaser_subspace_norm": 64.88258361816406, "collisions_per_episode": 0.550000011920929, "entropy": 0.046892862766981125, "explorer_grad_norm": 0.5321125462651253, "explorer_new_fields": 66.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 49.900001525878906, "explorer_subspace_norm": 37.78989791870117, "final_distance": 4.708261013031006, "max_param_abs": 0.9060558080673218, "policy_loss": -0.0008475810824886132, "value_loss": 3.7141629457473755}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:53:16.789940+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989596.7899528, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.210243896945839, "elapsed_seconds": 4356.87802900502, "env_steps_per_second": 10558.018767972033, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8945280977425766, "eta_seconds": 3220.3011518732756, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 2.930658130950725e-06, "chaser_grad_norm": 0.05640668235719204, "chaser_new_fields": 82.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.425000190734863, "chaser_subspace_norm": 65.13447570800781, "collisions_per_episode": 1.125, "entropy": 0.06048096064478159, "explorer_grad_norm": 0.7052809670567513, "explorer_new_fields": 58.85000228881836, "explorer_partner_vision": 0.0, "explorer_return": 38.087501525878906, "explorer_subspace_norm": 42.135154724121094, "final_distance": 4.785346508026123, "max_param_abs": 0.9292958378791809, "policy_loss": -0.0009047697975788083, "value_loss": 4.1310728788375854}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:56:26.332378+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989786.3323889, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.263585625274443, "elapsed_seconds": 4548.908250987995, "env_steps_per_second": 10551.982443166376, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8423904168496289, "eta_seconds": 3032.605500658664, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0004694193742231789, "chaser_grad_norm": 0.12446961924433708, "chaser_new_fields": 82.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.465001106262207, "chaser_subspace_norm": 59.04104232788086, "collisions_per_episode": 0.824999988079071, "entropy": 0.0753172505646944, "explorer_grad_norm": 1.732628419995308, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.0, "explorer_return": 37.45000076293945, "explorer_subspace_norm": 39.9908332824707, "final_distance": 4.435633182525635, "max_param_abs": 1.0153135061264038, "policy_loss": -0.001206845864420103, "value_loss": 3.9821478128433228}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:59:38.362602+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989978.362618, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3160478826791708, "elapsed_seconds": 4737.772377645015, "env_steps_per_second": 10553.482948214853, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7896287296075024, "eta_seconds": 2842.663426587009, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005332372871447166, "chaser_grad_norm": 0.07751051336526871, "chaser_new_fields": 84.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.770001411437988, "chaser_subspace_norm": 64.60179138183594, "collisions_per_episode": 0.8500000238418579, "entropy": 0.07866857387125492, "explorer_grad_norm": 1.1546983122825623, "explorer_new_fields": 61.54999923706055, "explorer_partner_vision": 0.0, "explorer_return": 42.025001525878906, "explorer_subspace_norm": 42.076744079589844, "final_distance": 4.865266799926758, "max_param_abs": 1.0153135061264038, "policy_loss": -0.0008040200193732971, "value_loss": 3.7884740829467773}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:02:47.226726+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990167.2267373, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3685334012750212, "elapsed_seconds": 4926.720244590077, "env_steps_per_second": 10554.689005753891, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7369026006865499, "eta_seconds": 2652.8493624715798, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0003707008744518736, "chaser_grad_norm": 0.0572335422039032, "chaser_new_fields": 80.125, "chaser_partner_vision": 0.0, "chaser_return": 5.935000896453857, "chaser_subspace_norm": 66.55184173583984, "collisions_per_episode": 1.0500000715255737, "entropy": 0.09421567432582378, "explorer_grad_norm": 0.7702902406454086, "explorer_new_fields": 63.22500228881836, "explorer_partner_vision": 0.0, "explorer_return": 44.53750228881836, "explorer_subspace_norm": 43.0537109375, "final_distance": 5.305086612701416, "max_param_abs": 1.0135619640350342, "policy_loss": -0.0008524347126979137, "value_loss": 4.081601023674011}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:05:56.174597+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990356.1746116, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4207160516311221, "elapsed_seconds": 5114.577785872039, "env_steps_per_second": 10558.056258165396, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6840484693038735, "eta_seconds": 2462.574489493945, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0002507129831073873, "chaser_grad_norm": 0.09457024000585079, "chaser_new_fields": 83.375, "chaser_partner_vision": 0.0, "chaser_return": 6.580001354217529, "chaser_subspace_norm": 69.01710510253906, "collisions_per_episode": 1.350000023841858, "entropy": 0.06742289289832115, "explorer_grad_norm": 0.49649740755558014, "explorer_new_fields": 63.150001525878906, "explorer_partner_vision": 0.0, "explorer_return": 44.23749923706055, "explorer_subspace_norm": 40.59196853637695, "final_distance": 4.132762908935547, "max_param_abs": 1.0311267375946045, "policy_loss": -0.00034488630975459955, "value_loss": 3.8211877942085266}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:09:04.032135+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990544.0321465, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4743230689561138, "elapsed_seconds": 5307.563048242009, "env_steps_per_second": 10550.981588160037, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6318527438383345, "eta_seconds": 2274.669877818004, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0005674249387863095, "chaser_grad_norm": 0.06129732821136713, "chaser_new_fields": 81.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.1750006675720215, "chaser_subspace_norm": 69.92224884033203, "collisions_per_episode": 1.975000023841858, "entropy": 0.10159408487379551, "explorer_grad_norm": 0.2755318284034729, "explorer_new_fields": 61.125, "explorer_partner_vision": 0.0, "explorer_return": 40.9375, "explorer_subspace_norm": 41.24720764160156, "final_distance": 4.750110626220703, "max_param_abs": 1.041700005531311, "policy_loss": -0.0009989746959178092, "value_loss": 4.173528671264648}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:12:17.017398+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990737.017413, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5263042088550154, "elapsed_seconds": 5494.695151878055, "env_steps_per_second": 10555.635644349793, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.578942975772592, "eta_seconds": 2084.1947127813314, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0004207111673258934, "chaser_grad_norm": 0.1405296865850687, "chaser_new_fields": 80.625, "chaser_partner_vision": 0.0, "chaser_return": 6.045001029968262, "chaser_subspace_norm": 67.5808334350586, "collisions_per_episode": 1.5500000715255737, "entropy": 0.06873511523008347, "explorer_grad_norm": 0.4632001221179962, "explorer_new_fields": 60.60000228881836, "explorer_partner_vision": 0.0, "explorer_return": 40.45000076293945, "explorer_subspace_norm": 43.44976043701172, "final_distance": 4.702489376068115, "max_param_abs": 1.1101652383804321, "policy_loss": -0.00019227069625316062, "value_loss": 3.984876334667206}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:15:24.149500+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990924.149512, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5782684272197205, "elapsed_seconds": 5681.766337990994, "env_steps_per_second": 10560.096355742658, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5260894757399068, "eta_seconds": 1893.9221126636644, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004060883537935697, "chaser_grad_norm": 0.07469570823013783, "chaser_new_fields": 81.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.250000953674316, "chaser_subspace_norm": 71.8243408203125, "collisions_per_episode": 1.149999976158142, "entropy": 0.0646466612815857, "explorer_grad_norm": 0.7807796597480774, "explorer_new_fields": 59.47500228881836, "explorer_partner_vision": 0.0, "explorer_return": 38.6875, "explorer_subspace_norm": 42.91965103149414, "final_distance": 4.014322757720947, "max_param_abs": 1.1201974153518677, "policy_loss": -0.0004118931193346498, "value_loss": 3.674412429332733}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:18:31.220689+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991111.220703, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6311393408791628, "elapsed_seconds": 5872.101627164986, "env_steps_per_second": 10558.40037120291, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4735565828358859, "eta_seconds": 1704.8036982091894, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 1.9749080453740353e-05, "chaser_grad_norm": 0.0697549618780613, "chaser_new_fields": 80.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.110001087188721, "chaser_subspace_norm": 70.33245849609375, "collisions_per_episode": 1.274999976158142, "entropy": 0.07250789552927017, "explorer_grad_norm": 0.7707884833216667, "explorer_new_fields": 61.150001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.3125, "explorer_subspace_norm": 43.19846725463867, "final_distance": 3.808316946029663, "max_param_abs": 1.131509780883789, "policy_loss": -0.000504695579174097, "value_loss": 3.7307079434394836}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:21:41.555976+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991301.5559874, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6838425399780519, "elapsed_seconds": 6061.833143920987, "env_steps_per_second": 10557.86236283679, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.42096063499451297, "eta_seconds": 1515.4582859802467, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00031742011053230434, "chaser_grad_norm": 0.07916883938014507, "chaser_new_fields": 78.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.625000953674316, "chaser_subspace_norm": 68.92870330810547, "collisions_per_episode": 1.9250000715255737, "entropy": 0.05154225882142782, "explorer_grad_norm": 0.28872742503881454, "explorer_new_fields": 62.72500228881836, "explorer_partner_vision": 0.0, "explorer_return": 43.41250228881836, "explorer_subspace_norm": 46.36096954345703, "final_distance": 4.864537715911865, "max_param_abs": 1.151167869567871, "policy_loss": -0.000658438459504751, "value_loss": 3.887501835823059}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:24:51.287493+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991491.2875078, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7365668448111198, "elapsed_seconds": 6251.640641320031, "env_steps_per_second": 10557.228699899188, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3683626640508436, "eta_seconds": 1326.105590583037, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": -1.4677507732763218e-05, "chaser_grad_norm": 0.09247189201414585, "chaser_new_fields": 80.25, "chaser_partner_vision": 0.0, "chaser_return": 5.995000839233398, "chaser_subspace_norm": 75.86627960205078, "collisions_per_episode": 1.2000000476837158, "entropy": 0.04299487080425024, "explorer_grad_norm": 0.5087362602353096, "explorer_new_fields": 60.67499923706055, "explorer_partner_vision": 0.0, "explorer_return": 40.63750076293945, "explorer_subspace_norm": 46.21202850341797, "final_distance": 4.302360534667969, "max_param_abs": 1.1721464395523071, "policy_loss": -0.00045451677363672616, "value_loss": 3.6913729310035706}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:28:01.094990+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991681.0950007, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7889445060427858, "elapsed_seconds": 6440.2002217540285, "env_steps_per_second": 10558.677938351391, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3156960893016681, "eta_seconds": 1136.505921486005, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0006146005711815672, "chaser_grad_norm": 0.03832845529541373, "chaser_new_fields": 79.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.915000915527344, "chaser_subspace_norm": 67.42417907714844, "collisions_per_episode": 1.125, "entropy": 0.07507002167403698, "explorer_grad_norm": 0.9272700697183609, "explorer_new_fields": 60.92499923706055, "explorer_partner_vision": 0.0, "explorer_return": 40.75, "explorer_subspace_norm": 44.226131439208984, "final_distance": 4.725557804107666, "max_param_abs": 1.2197116613388062, "policy_loss": -0.0007195406943791838, "value_loss": 3.8805500268936157}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:31:09.654573+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991869.6545868, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8427941666197472, "elapsed_seconds": 6634.05899983109, "env_steps_per_second": 10551.609505098202, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26325630951710677, "eta_seconds": 947.7227142615843, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00018568683959248322, "chaser_grad_norm": 0.05355698801577091, "chaser_new_fields": 81.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.135000705718994, "chaser_subspace_norm": 75.23600769042969, "collisions_per_episode": 0.949999988079071, "entropy": 0.06457692570984364, "explorer_grad_norm": 0.7618395239114761, "explorer_new_fields": 61.275001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.35000228881836, "explorer_subspace_norm": 44.906795501708984, "final_distance": 4.91195821762085, "max_param_abs": 1.216644048690796, "policy_loss": -0.0004897058455988201, "value_loss": 3.892449378967285}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:34:23.513351+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992063.5133636, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.895706312488053, "elapsed_seconds": 6824.542724956991, "env_steps_per_second": 10550.157410063508, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21063403472089476, "eta_seconds": 758.2825249952211, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0004463492454132023, "chaser_grad_norm": 0.08135259244590998, "chaser_new_fields": 80.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.095001220703125, "chaser_subspace_norm": 73.42074584960938, "collisions_per_episode": 0.925000011920929, "entropy": 0.08478923514485359, "explorer_grad_norm": 1.1862336099147797, "explorer_new_fields": 62.25, "explorer_partner_vision": 0.0, "explorer_return": 42.92499923706055, "explorer_subspace_norm": 43.7446174621582, "final_distance": 4.825169563293457, "max_param_abs": 1.2055058479309082, "policy_loss": -0.0006592626589139172, "value_loss": 3.8866884112358093}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:37:33.997074+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992253.997089, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9488581179416764, "elapsed_seconds": 7015.889224590035, "env_steps_per_second": 10547.486944439905, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15801552307635214, "eta_seconds": 568.8558830748677, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0001439858323394501, "chaser_grad_norm": 0.08166112005710602, "chaser_new_fields": 80.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.0550007820129395, "chaser_subspace_norm": 81.14865112304688, "collisions_per_episode": 1.25, "entropy": 0.06311777606606483, "explorer_grad_norm": 1.4002510160207748, "explorer_new_fields": 62.29999923706055, "explorer_partner_vision": 0.0, "explorer_return": 42.962501525878906, "explorer_subspace_norm": 47.11442184448242, "final_distance": 4.7119550704956055, "max_param_abs": 1.161903977394104, "policy_loss": -0.00023089893331107447, "value_loss": 3.708661198616028}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:40:45.343575+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992445.3435903, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.001409211811681, "elapsed_seconds": 7205.073162522051, "env_steps_per_second": 10548.123285592994, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1053373269374569, "eta_seconds": 379.2143769748448, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00040755003191149386, "chaser_grad_norm": 0.0962408883497119, "chaser_new_fields": 82.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.395000457763672, "chaser_subspace_norm": 79.08171081542969, "collisions_per_episode": 1.5750000476837158, "entropy": 0.055252641439437866, "explorer_grad_norm": 0.9006635546684265, "explorer_new_fields": 61.47500228881836, "explorer_partner_vision": 0.0, "explorer_return": 41.650001525878906, "explorer_subspace_norm": 47.367366790771484, "final_distance": 4.663539409637451, "max_param_abs": 1.1662650108337402, "policy_loss": -0.0008348307880936545, "value_loss": 3.6246113181114197}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:43:54.527513+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992634.527528, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.053409373241108, "elapsed_seconds": 7392.273743667989, "env_steps_per_second": 10551.557302218762, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05265152239079765, "eta_seconds": 189.54548060687154, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005645156096885096, "chaser_grad_norm": 0.13609989918768406, "chaser_new_fields": 81.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.140000820159912, "chaser_subspace_norm": 77.23444366455078, "collisions_per_episode": 1.649999976158142, "entropy": 0.06489465571939945, "explorer_grad_norm": 0.7174627259373665, "explorer_new_fields": 62.22500228881836, "explorer_partner_vision": 0.0, "explorer_return": 42.66250228881836, "explorer_subspace_norm": 50.78065490722656, "final_distance": 4.717092037200928, "max_param_abs": 1.210858941078186, "policy_loss": -0.0010156546933615118, "value_loss": 3.939854860305786}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:47:01.728093+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992821.7281044, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1019971768872345, "elapsed_seconds": 7567.189836794045, "env_steps_per_second": 10571.956264532306, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:49:56.644185+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992996.6441965, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1019998172186187, "elapsed_seconds": 7567.199341987027, "env_steps_per_second": 10571.942985050697, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:49:56.653691+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992996.6537013, "update": 20000} diff --git a/non_social/seed_0006/attempt_01/status.json b/non_social/seed_0006/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..616b86268da142480fbf79f341497e60c2317ad5 --- /dev/null +++ b/non_social/seed_0006/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1019998172186187, + "elapsed_seconds": 7567.199341987027, + "env_steps_per_second": 10571.942985050697, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 8.635803430578903e-06, + "chaser_grad_norm": 0.10996681079268456, + "chaser_new_fields": 82.375, + "chaser_partner_vision": 0.0, + "chaser_return": 6.390000820159912, + "chaser_subspace_norm": 73.99776458740234, + "collisions_per_episode": 1.149999976158142, + "entropy": 0.05159799661487341, + "explorer_grad_norm": 0.7565127648413181, + "explorer_new_fields": 61.025001525878906, + "explorer_partner_vision": 0.0, + "explorer_return": 41.23749923706055, + "explorer_subspace_norm": 46.546077728271484, + "final_distance": 4.668344020843506, + "max_param_abs": 1.2401388883590698, + "policy_loss": -0.0002673804239634592, + "value_loss": 3.8410890698432922 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:49:56.653691+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782992996.6537013, + "update": 20000 +} diff --git a/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d2216131f036cd6a2b7d93f70d97f1d961735ff --- /dev/null +++ b/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:556bd8154b82d372e4f881e6319eeee0a1ad15f4e6c1b04df358d0711c48283b +size 11460392 diff --git a/non_social/seed_0007/attempt_01/config.json b/non_social/seed_0007/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..5ce385424fb706ef60efd09b450d4376ee82d030 --- /dev/null +++ b/non_social/seed_0007/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", + "seed": 7, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0007/attempt_01/metrics.jsonl b/non_social/seed_0007/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..7d502056a175b25482277bd0d60b38fcec5e7ef5 --- /dev/null +++ b/non_social/seed_0007/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006162312338710762, "elapsed_seconds": 2.2184324419358745, "env_steps_per_second": 1803.07496608257, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 12.324008446187655, "eta_seconds": 44366.430406275555, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 1.5863511957356358e-05, "chaser_grad_norm": 0.40252913534641266, "chaser_new_fields": 34.125, "chaser_partner_vision": 0.0, "chaser_return": -3.2400002479553223, "chaser_subspace_norm": 0.0026733726263046265, "collisions_per_episode": 1.475000023841858, "entropy": 2.772510290145874, "explorer_grad_norm": 0.1284720953553915, "explorer_new_fields": 34.54999923706055, "explorer_partner_vision": 0.0, "explorer_return": 1.4875000715255737, "explorer_subspace_norm": 0.002178831957280636, "final_distance": 5.129796504974365, "max_param_abs": 0.15428495407104492, "policy_loss": -0.00041444065969020727, "value_loss": 3.1294726133346558}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:44:02.071502+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985442.0715215, "update": 1} +{"device": "cuda", "elapsed_hours": 0.0534089812913832, "elapsed_seconds": 192.2723326489795, "env_steps_per_second": 10401.912602013752, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0829502703639444, "eta_seconds": 7498.6209733102005, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0007816769169011195, "chaser_grad_norm": 0.23463235050439835, "chaser_new_fields": 74.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 4.6650004386901855, "chaser_subspace_norm": 16.046295166015625, "collisions_per_episode": 3.0250000953674316, "entropy": 0.433866485953331, "explorer_grad_norm": 0.4533975422382355, "explorer_new_fields": 63.97500228881836, "explorer_partner_vision": 0.0, "explorer_return": 45.025001525878906, "explorer_subspace_norm": 9.584186553955078, "final_distance": 5.005619525909424, "max_param_abs": 0.32394322752952576, "policy_loss": -0.0023453493630505307, "value_loss": 4.62317430973053}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:47:12.125401+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985632.1254144, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10704267039471435, "elapsed_seconds": 385.35361342097167, "env_steps_per_second": 10380.076534095664, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.033810737499573, "eta_seconds": 7321.718654998462, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00030918466840529346, "chaser_grad_norm": 0.3375948593020439, "chaser_new_fields": 75.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.9900007247924805, "chaser_subspace_norm": 27.383764266967773, "collisions_per_episode": 2.875, "entropy": 0.24814743176102638, "explorer_grad_norm": 0.5768401697278023, "explorer_new_fields": 64.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 46.22500228881836, "explorer_subspace_norm": 12.494498252868652, "final_distance": 5.348890781402588, "max_param_abs": 0.3940936326980591, "policy_loss": -0.0014165264188772397, "value_loss": 4.558527231216431}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:50:25.206680+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985825.2066953, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16053553507081234, "elapsed_seconds": 577.9279262549244, "env_steps_per_second": 10381.917411191851, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9799382658733524, "eta_seconds": 7127.777757144068, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 8.75314920669723e-05, "chaser_grad_norm": 0.15555922500789165, "chaser_new_fields": 79.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.850000858306885, "chaser_subspace_norm": 35.83333969116211, "collisions_per_episode": 2.5250000953674316, "entropy": 0.15090478584170341, "explorer_grad_norm": 0.9521286189556122, "explorer_new_fields": 68.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 51.25, "explorer_subspace_norm": 15.684005737304688, "final_distance": 4.942183971405029, "max_param_abs": 0.4130028784275055, "policy_loss": -0.0011707026369407458, "value_loss": 4.426733493804932}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:37.780991+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986017.7810025, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21307040575387268, "elapsed_seconds": 767.0534607139416, "env_steps_per_second": 10429.52076971784, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9176336517848538, "eta_seconds": 6903.4811464254735, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00012393530064658265, "chaser_grad_norm": 0.1781393364071846, "chaser_new_fields": 79.375, "chaser_partner_vision": 0.0, "chaser_return": 5.680001258850098, "chaser_subspace_norm": 39.175682067871094, "collisions_per_episode": 2.875, "entropy": 0.09888806566596031, "explorer_grad_norm": 0.2888704016804695, "explorer_new_fields": 71.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 56.42499923706055, "explorer_subspace_norm": 18.411497116088867, "final_distance": 6.109438896179199, "max_param_abs": 0.4200720489025116, "policy_loss": -0.0008519365746764151, "value_loss": 4.206562280654907}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:46.906527+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986206.9065409, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2647814488644427, "elapsed_seconds": 953.2132159119938, "env_steps_per_second": 10490.832306004513, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.853470142051099, "eta_seconds": 6672.492511383956, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 8.50081906200173e-05, "chaser_grad_norm": 0.1132832020521164, "chaser_new_fields": 79.75, "chaser_partner_vision": 0.0, "chaser_return": 5.870000839233398, "chaser_subspace_norm": 46.261566162109375, "collisions_per_episode": 1.875, "entropy": 0.0645553432404995, "explorer_grad_norm": 1.0586529970169067, "explorer_new_fields": 74.75, "explorer_partner_vision": 0.0, "explorer_return": 61.375, "explorer_subspace_norm": 20.723173141479492, "final_distance": 4.829738140106201, "max_param_abs": 0.4422103464603424, "policy_loss": -0.0006739977858379875, "value_loss": 3.954309582710266}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:53.066281+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986393.066292, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3176759706702756, "elapsed_seconds": 1143.6334944129921, "env_steps_per_second": 10492.872112109131, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8001638337982286, "eta_seconds": 6480.589801673623, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00010335987570288252, "chaser_grad_norm": 0.05535600148141384, "chaser_new_fields": 75.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 4.885000705718994, "chaser_subspace_norm": 50.445125579833984, "collisions_per_episode": 2.4000000953674316, "entropy": 0.07440762594342232, "explorer_grad_norm": 0.6606214866042137, "explorer_new_fields": 75.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.53750228881836, "explorer_subspace_norm": 24.411928176879883, "final_distance": 4.508893013000488, "max_param_abs": 0.49596044421195984, "policy_loss": -0.0004586243771456111, "value_loss": 4.105130434036255}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:03:03.486560+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986583.4865735, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3707698364844287, "elapsed_seconds": 1334.7714113439433, "env_steps_per_second": 10488.687337035335, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7479149434265924, "eta_seconds": 6292.493796335732, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0004017523753252483, "chaser_grad_norm": 0.24306313693523407, "chaser_new_fields": 81.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.160000801086426, "chaser_subspace_norm": 53.54353332519531, "collisions_per_episode": 2.424999952316284, "entropy": 0.06470314040780067, "explorer_grad_norm": 1.4010419249534607, "explorer_new_fields": 73.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 59.38750076293945, "explorer_subspace_norm": 26.491323471069336, "final_distance": 5.1734747886657715, "max_param_abs": 0.7580186724662781, "policy_loss": -0.0008832078501512797, "value_loss": 3.992945671081543}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:06:14.624476+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986774.6244872, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4227425162000064, "elapsed_seconds": 1521.8730583200231, "env_steps_per_second": 10513.360436028877, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6909700648000257, "eta_seconds": 6087.4922332800925, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00020032795533662245, "chaser_grad_norm": 0.15399200841784477, "chaser_new_fields": 80.125, "chaser_partner_vision": 0.0, "chaser_return": 5.890000820159912, "chaser_subspace_norm": 61.421199798583984, "collisions_per_episode": 2.549999952316284, "entropy": 0.0632891058921814, "explorer_grad_norm": 0.7257499247789383, "explorer_new_fields": 77.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 65.6500015258789, "explorer_subspace_norm": 28.03260612487793, "final_distance": 5.098093509674072, "max_param_abs": 0.8237745761871338, "policy_loss": -0.000726720602891362, "value_loss": 3.974096417427063}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:09:21.726124+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986961.7261374, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.47565659327972554, "elapsed_seconds": 1712.3637358070118, "env_steps_per_second": 10511.78533135477, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.638372710185721, "eta_seconds": 5898.141756668596, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 1.9615165482239627e-05, "chaser_grad_norm": 0.12753227539360523, "chaser_new_fields": 76.5, "chaser_partner_vision": 0.0, "chaser_return": 5.195000648498535, "chaser_subspace_norm": 59.81772994995117, "collisions_per_episode": 2.549999952316284, "entropy": 0.047921315766870975, "explorer_grad_norm": 0.6383462622761726, "explorer_new_fields": 75.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 61.82500076293945, "explorer_subspace_norm": 29.96605682373047, "final_distance": 5.032867908477783, "max_param_abs": 0.9751938581466675, "policy_loss": -0.00038563635057542456, "value_loss": 4.066440463066101}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:32.216801+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987152.2168128, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5296401801505514, "elapsed_seconds": 1906.704648541985, "env_steps_per_second": 10489.30153670814, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5889205404516542, "eta_seconds": 5720.113945625955, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005427189815506404, "chaser_grad_norm": 0.10228882543742657, "chaser_new_fields": 74.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.790000915527344, "chaser_subspace_norm": 63.0869140625, "collisions_per_episode": 2.2750000953674316, "entropy": 0.07300866022706032, "explorer_grad_norm": 0.9975150376558304, "explorer_new_fields": 76.75, "explorer_partner_vision": 0.0, "explorer_return": 64.5250015258789, "explorer_subspace_norm": 31.419540405273438, "final_distance": 5.290417671203613, "max_param_abs": 1.0179098844528198, "policy_loss": -0.0010624551471494215, "value_loss": 3.9885209798812866}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:46.557715+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987346.557729, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5830853869241481, "elapsed_seconds": 2099.1073929269332, "env_steps_per_second": 10480.645284815015, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5372251109818449, "eta_seconds": 5534.010399534642, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": -2.2846327839581848e-05, "chaser_grad_norm": 0.06350257247686386, "chaser_new_fields": 78.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.460000514984131, "chaser_subspace_norm": 77.6115493774414, "collisions_per_episode": 2.700000047683716, "entropy": 0.03230392653495073, "explorer_grad_norm": 0.4880380630493164, "explorer_new_fields": 79.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 68.1624984741211, "explorer_subspace_norm": 32.83280563354492, "final_distance": 4.66172981262207, "max_param_abs": 1.1070725917816162, "policy_loss": -0.0007277827493403777, "value_loss": 3.926747679710388}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:18:58.960460+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987538.9604733, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6360462296033317, "elapsed_seconds": 2289.766426571994, "env_steps_per_second": 10481.418419576692, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4841078690744407, "eta_seconds": 5342.788328667986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00043554698165632066, "chaser_grad_norm": 0.08591583743691444, "chaser_new_fields": 80.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.890000820159912, "chaser_subspace_norm": 72.68104553222656, "collisions_per_episode": 2.25, "entropy": 0.028174152597784996, "explorer_grad_norm": 0.48343631625175476, "explorer_new_fields": 77.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 65.9124984741211, "explorer_subspace_norm": 33.87345504760742, "final_distance": 4.6805853843688965, "max_param_abs": 1.1083637475967407, "policy_loss": -0.0008415480105909978, "value_loss": 3.83197283744812}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:22:09.619492+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987729.6195078, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6889137845877688, "elapsed_seconds": 2480.0896245159674, "env_steps_per_second": 10483.492105683217, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4308209372207503, "eta_seconds": 5150.955373994701, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00033065811358401476, "chaser_grad_norm": 0.13097486458718777, "chaser_new_fields": 76.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.1750006675720215, "chaser_subspace_norm": 65.27603149414062, "collisions_per_episode": 2.325000047683716, "entropy": 0.056820694357156754, "explorer_grad_norm": 0.42283712327480316, "explorer_new_fields": 76.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 63.92499923706055, "explorer_subspace_norm": 34.83368682861328, "final_distance": 4.935168743133545, "max_param_abs": 1.1083637475967407, "policy_loss": -0.0006907511765601448, "value_loss": 4.0870561599731445}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:25:19.942689+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987919.9427001, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7419893949388643, "elapsed_seconds": 2671.1618217799114, "env_steps_per_second": 10482.330112573405, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.377980304886462, "eta_seconds": 4960.729097591264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0005611336601397954, "chaser_grad_norm": 0.36040082573890686, "chaser_new_fields": 76.375, "chaser_partner_vision": 0.0, "chaser_return": 5.180001258850098, "chaser_subspace_norm": 70.00305938720703, "collisions_per_episode": 2.174999952316284, "entropy": 0.06030638702213764, "explorer_grad_norm": 0.5832213535904884, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.2125015258789, "explorer_subspace_norm": 34.063011169433594, "final_distance": 4.883917331695557, "max_param_abs": 1.1135412454605103, "policy_loss": -0.0013945258434446828, "value_loss": 4.070422530174255}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:28:31.014887+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988111.0149002, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.793640891656105, "elapsed_seconds": 2857.107209961978, "env_steps_per_second": 10500.131005024217, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3227348194268416, "eta_seconds": 4761.84534993663, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0008002171707049005, "chaser_grad_norm": 0.10145391151309013, "chaser_new_fields": 78.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.5450005531311035, "chaser_subspace_norm": 73.43029022216797, "collisions_per_episode": 4.025000095367432, "entropy": 0.08901901729404926, "explorer_grad_norm": 0.5299194008111954, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.01250076293945, "explorer_subspace_norm": 33.322715759277344, "final_distance": 5.0767083168029785, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0014485676968476469, "value_loss": 3.9985559582710266}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:31:36.960274+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988296.9602861, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8470204308591525, "elapsed_seconds": 3049.2735510929488, "env_steps_per_second": 10494.302811412334, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2705306462887287, "eta_seconds": 4573.910326639423, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00018901966278606075, "chaser_grad_norm": 0.08398678712546825, "chaser_new_fields": 80.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.835000514984131, "chaser_subspace_norm": 80.05250549316406, "collisions_per_episode": 2.625, "entropy": 0.054006004706025124, "explorer_grad_norm": 0.3982117027044296, "explorer_new_fields": 79.125, "explorer_partner_vision": 0.0, "explorer_return": 68.0875015258789, "explorer_subspace_norm": 33.364192962646484, "final_distance": 4.895865440368652, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0010097663823955827, "value_loss": 3.6410703659057617}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:34:49.126619+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988489.1266334, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9002772220599905, "elapsed_seconds": 3240.9979994159658, "env_steps_per_second": 10490.595799851424, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2180221239635165, "eta_seconds": 4384.87964626866, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 9.515890963940343e-05, "chaser_grad_norm": 0.08019938878715038, "chaser_new_fields": 83.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.575000762939453, "chaser_subspace_norm": 78.6480712890625, "collisions_per_episode": 2.6000001430511475, "entropy": 0.060854741372168064, "explorer_grad_norm": 0.5011994987726212, "explorer_new_fields": 76.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 63.625, "explorer_subspace_norm": 32.06963348388672, "final_distance": 5.302577018737793, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0006163118757536168, "value_loss": 3.7890986800193787}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:38:00.851065+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988680.8510792, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9539020866588948, "elapsed_seconds": 3434.0475119720213, "env_steps_per_second": 10483.256237572205, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1658803281386492, "eta_seconds": 4197.169181299137, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0002007880757145114, "chaser_grad_norm": 0.0504648070782423, "chaser_new_fields": 80.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.990001201629639, "chaser_subspace_norm": 74.63182830810547, "collisions_per_episode": 2.049999952316284, "entropy": 0.066467409953475, "explorer_grad_norm": 1.3335921242833138, "explorer_new_fields": 74.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 60.66250228881836, "explorer_subspace_norm": 32.54383850097656, "final_distance": 4.8907670974731445, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0006288228665383144, "value_loss": 3.98265278339386}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:41:13.900578+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988873.9005928, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0056183147388704, "elapsed_seconds": 3620.2259330599336, "env_steps_per_second": 10496.582451659628, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1114728741850672, "eta_seconds": 4001.3023470662424, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 7.321380738944949e-05, "chaser_grad_norm": 0.15117140114307404, "chaser_new_fields": 84.5, "chaser_partner_vision": 0.0, "chaser_return": 6.7350006103515625, "chaser_subspace_norm": 86.46621704101562, "collisions_per_episode": 2.4750001430511475, "entropy": 0.040792311541736126, "explorer_grad_norm": 1.2298443838953972, "explorer_new_fields": 75.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 62.462501525878906, "explorer_subspace_norm": 34.72623825073242, "final_distance": 4.450304985046387, "max_param_abs": 1.1041847467422485, "policy_loss": -0.000505954831723443, "value_loss": 3.622510254383087}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:44:20.078997+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989060.0790102, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0578598396830299, "elapsed_seconds": 3808.295422858908, "env_steps_per_second": 10503.386832834462, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0578598396830299, "eta_seconds": 3808.295422858908, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 5.705474892581108e-05, "chaser_grad_norm": 0.33392541110515594, "chaser_new_fields": 79.125, "chaser_partner_vision": 0.0, "chaser_return": 5.660000801086426, "chaser_subspace_norm": 88.33106231689453, "collisions_per_episode": 3.200000047683716, "entropy": 0.04662791267037392, "explorer_grad_norm": 1.1790385693311691, "explorer_new_fields": 74.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 61.1875, "explorer_subspace_norm": 34.119956970214844, "final_distance": 4.180972576141357, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0007274195271342876, "value_loss": 3.959327459335327}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:47:28.148488+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989248.1485012, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1099273963508312, "elapsed_seconds": 3995.7386268629925, "env_steps_per_second": 10511.198034235214, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0042200252697997, "eta_seconds": 3615.192090971279, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -5.344487079478322e-06, "chaser_grad_norm": 0.14296459965407848, "chaser_new_fields": 80.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.0500006675720215, "chaser_subspace_norm": 84.44515228271484, "collisions_per_episode": 2.75, "entropy": 0.0408216267824173, "explorer_grad_norm": 0.7078053206205368, "explorer_new_fields": 74.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 61.22500228881836, "explorer_subspace_norm": 33.62204360961914, "final_distance": 4.204414367675781, "max_param_abs": 1.1041847467422485, "policy_loss": -0.00045505150721858456, "value_loss": 3.824666380882263}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:35.591691+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989435.591703, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1625295734924859, "elapsed_seconds": 4185.106464572949, "env_steps_per_second": 10513.472087857575, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9511605601302158, "eta_seconds": 3424.1780164687766, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0001621657335137, "chaser_grad_norm": 0.05781709309667349, "chaser_new_fields": 83.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 96.06954956054688, "collisions_per_episode": 2.4750001430511475, "entropy": 0.04814101289957762, "explorer_grad_norm": 0.6472741663455963, "explorer_new_fields": 72.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 58.5625, "explorer_subspace_norm": 33.06576156616211, "final_distance": 4.316286563873291, "max_param_abs": 1.1041847467422485, "policy_loss": -0.000244601264014932, "value_loss": 3.884456217288971}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:53:44.959531+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989624.9595451, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.214664357043059, "elapsed_seconds": 4372.791685355012, "env_steps_per_second": 10519.595560442394, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8977953943361741, "eta_seconds": 3232.0634196102264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00019810608695618181, "chaser_grad_norm": 0.27932218462228775, "chaser_new_fields": 81.5, "chaser_partner_vision": 0.0, "chaser_return": 6.155001163482666, "chaser_subspace_norm": 94.1336441040039, "collisions_per_episode": 2.625, "entropy": 0.031551075633615255, "explorer_grad_norm": 1.2909544557332993, "explorer_new_fields": 78.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 66.8125, "explorer_subspace_norm": 34.45367431640625, "final_distance": 4.879254341125488, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0004625879838133784, "value_loss": 3.7053160071372986}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:56:52.644752+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989812.6447659, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2681079213299866, "elapsed_seconds": 4565.188516787952, "env_steps_per_second": 10514.352216449674, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8454052808866578, "eta_seconds": 3043.459011191968, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005135875747448969, "chaser_grad_norm": 0.052880093455314636, "chaser_new_fields": 83.0, "chaser_partner_vision": 0.0, "chaser_return": 6.465001106262207, "chaser_subspace_norm": 97.11492919921875, "collisions_per_episode": 2.625, "entropy": 0.03495701961219311, "explorer_grad_norm": 0.5557614490389824, "explorer_new_fields": 78.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 67.75, "explorer_subspace_norm": 34.69092559814453, "final_distance": 5.4152326583862305, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0007133961352519536, "value_loss": 3.71113383769989}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:00:05.041583+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990005.0415983, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.321793889427758, "elapsed_seconds": 4758.458001939929, "env_steps_per_second": 10507.605610812576, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7930763336566549, "eta_seconds": 2855.0748011639575, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00027742349917787845, "chaser_grad_norm": 0.2512429542839527, "chaser_new_fields": 79.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.840001106262207, "chaser_subspace_norm": 99.58148193359375, "collisions_per_episode": 2.2250001430511475, "entropy": 0.050363486632704735, "explorer_grad_norm": 0.36915355175733566, "explorer_new_fields": 75.0, "explorer_partner_vision": 0.0, "explorer_return": 61.900001525878906, "explorer_subspace_norm": 35.681427001953125, "final_distance": 5.007301330566406, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0009037921966488227, "value_loss": 3.897596061229706}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:03:18.311068+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990198.3110826, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3749829681138768, "elapsed_seconds": 4949.938685209956, "env_steps_per_second": 10505.18063089793, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7403754443690106, "eta_seconds": 2665.351599728438, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0003058196967000437, "chaser_grad_norm": 0.18610444478690624, "chaser_new_fields": 80.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.975001335144043, "chaser_subspace_norm": 94.67697143554688, "collisions_per_episode": 2.7250001430511475, "entropy": 0.053884525783360004, "explorer_grad_norm": 0.9429795071482658, "explorer_new_fields": 76.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 63.73749923706055, "explorer_subspace_norm": 33.789344787597656, "final_distance": 5.37762975692749, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0007647984459779877, "value_loss": 3.8410302996635437}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:06:29.791752+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990389.7917657, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.42673490026167, "elapsed_seconds": 5136.245640942012, "env_steps_per_second": 10513.515858656663, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6869464334593226, "eta_seconds": 2473.0071604535615, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 1.4911278931370453e-05, "chaser_grad_norm": 0.21987946704030037, "chaser_new_fields": 83.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.510000705718994, "chaser_subspace_norm": 89.98974609375, "collisions_per_episode": 2.049999952316284, "entropy": 0.02883374970406294, "explorer_grad_norm": 0.3827315643429756, "explorer_new_fields": 81.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 71.61250305175781, "explorer_subspace_norm": 35.02383041381836, "final_distance": 4.741418361663818, "max_param_abs": 1.1041847467422485, "policy_loss": -0.0003050914031348384, "value_loss": 3.6383501291275024}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:09:36.098705+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990576.0987177, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4804190003627768, "elapsed_seconds": 5329.508401305997, "env_steps_per_second": 10507.535739370856, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6344652858697615, "eta_seconds": 2284.0750291311415, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 9.075296678975886e-05, "chaser_grad_norm": 0.051150050945580006, "chaser_new_fields": 81.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.090001106262207, "chaser_subspace_norm": 91.92811584472656, "collisions_per_episode": 2.674999952316284, "entropy": 0.025759872514754534, "explorer_grad_norm": 1.0781443268060684, "explorer_new_fields": 83.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 74.01250457763672, "explorer_subspace_norm": 37.72677993774414, "final_distance": 4.748311996459961, "max_param_abs": 1.1305887699127197, "policy_loss": -0.00031955546583617433, "value_loss": 3.6773812770843506}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:12:49.361467+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990769.3614824, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5328695651077822, "elapsed_seconds": 5518.330434388015, "env_steps_per_second": 10510.42533418574, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.581433283316745, "eta_seconds": 2093.159819940282, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0003920660119590587, "chaser_grad_norm": 0.21297385171055794, "chaser_new_fields": 78.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.4850006103515625, "chaser_subspace_norm": 83.64649963378906, "collisions_per_episode": 3.2250001430511475, "entropy": 0.04538226220756769, "explorer_grad_norm": 1.09544737637043, "explorer_new_fields": 79.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 67.4124984741211, "explorer_subspace_norm": 36.844966888427734, "final_distance": 4.8579325675964355, "max_param_abs": 1.1544324159622192, "policy_loss": -0.0007786997561325393, "value_loss": 3.807976543903351}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:15:58.183499+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990958.1835103, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5864454033535893, "elapsed_seconds": 5711.203452072921, "env_steps_per_second": 10505.666713417919, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5288151344511964, "eta_seconds": 1903.7344840243072, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00019163531830918186, "chaser_grad_norm": 0.07847378961741924, "chaser_new_fields": 80.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.905001163482666, "chaser_subspace_norm": 92.05323791503906, "collisions_per_episode": 2.674999952316284, "entropy": 0.030779996421188116, "explorer_grad_norm": 0.5529287233948708, "explorer_new_fields": 83.0, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 39.154842376708984, "final_distance": 5.649028301239014, "max_param_abs": 1.1967204809188843, "policy_loss": -0.0006144566871206791, "value_loss": 3.792012333869934}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:19:11.056519+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991151.056533, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6398628940961013, "elapsed_seconds": 5903.506418745965, "env_steps_per_second": 10502.23301242216, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47608922731822295, "eta_seconds": 1713.9212183456027, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0003427448025421498, "chaser_grad_norm": 0.09134084871038795, "chaser_new_fields": 78.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.640000820159912, "chaser_subspace_norm": 94.06539916992188, "collisions_per_episode": 2.125, "entropy": 0.0441272733733058, "explorer_grad_norm": 0.428738534450531, "explorer_new_fields": 80.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 70.1500015258789, "explorer_subspace_norm": 37.67318344116211, "final_distance": 5.557610988616943, "max_param_abs": 1.2320512533187866, "policy_loss": -0.00018256950398942706, "value_loss": 3.8186123371124268}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:22:23.359485+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991343.3594987, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6935703084622058, "elapsed_seconds": 6096.8531104639405, "env_steps_per_second": 10497.218620891117, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.42339257711555145, "eta_seconds": 1524.2132776159851, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0003343662664766489, "chaser_grad_norm": 0.05997555796056986, "chaser_new_fields": 77.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.390000820159912, "chaser_subspace_norm": 103.74954986572266, "collisions_per_episode": 2.0, "entropy": 0.050705550238490105, "explorer_grad_norm": 0.393698550760746, "explorer_new_fields": 78.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 67.5250015258789, "explorer_subspace_norm": 37.62314224243164, "final_distance": 4.948698043823242, "max_param_abs": 1.2376559972763062, "policy_loss": -0.0004058454398503386, "value_loss": 3.8183892965316772}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:25:36.706176+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991536.7061896, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7452374496022174, "elapsed_seconds": 6282.854818567983, "env_steps_per_second": 10504.778783833655, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.37020188324895525, "eta_seconds": 1332.726779696239, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00027873318154159055, "chaser_grad_norm": 0.1305250469595194, "chaser_new_fields": 79.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.8000006675720215, "chaser_subspace_norm": 96.60128784179688, "collisions_per_episode": 2.6000001430511475, "entropy": 0.031662431079894304, "explorer_grad_norm": 0.3303103819489479, "explorer_new_fields": 80.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 69.73750305175781, "explorer_subspace_norm": 37.09413146972656, "final_distance": 5.164980411529541, "max_param_abs": 1.2631558179855347, "policy_loss": -0.000378172807919519, "value_loss": 3.7132776379585266}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:28:42.707883+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991722.707896, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7977941855952828, "elapsed_seconds": 6472.0590681430185, "env_steps_per_second": 10506.70262493614, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3172577974579911, "eta_seconds": 1142.128070848768, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 4.529805385337926e-05, "chaser_grad_norm": 0.15020860731601715, "chaser_new_fields": 82.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 91.23472595214844, "collisions_per_episode": 3.8500001430511475, "entropy": 0.07598509825766087, "explorer_grad_norm": 0.5112570002675056, "explorer_new_fields": 75.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 62.16250228881836, "explorer_subspace_norm": 34.94123458862305, "final_distance": 5.260217189788818, "max_param_abs": 1.287571668624878, "policy_loss": -0.0011185446090546947, "value_loss": 3.7376259565353394}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:31:51.912134+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991911.9121509, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8495232788294218, "elapsed_seconds": 6658.283803785918, "env_steps_per_second": 10513.219631791275, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26421761126134596, "eta_seconds": 951.1834005408455, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00018258933981496472, "chaser_grad_norm": 0.06957824435085058, "chaser_new_fields": 79.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.7950005531311035, "chaser_subspace_norm": 100.91751861572266, "collisions_per_episode": 2.75, "entropy": 0.060336608439683914, "explorer_grad_norm": 0.5515022277832031, "explorer_new_fields": 78.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 67.45000457763672, "explorer_subspace_norm": 36.68427276611328, "final_distance": 4.743283748626709, "max_param_abs": 1.3183913230895996, "policy_loss": -0.0005237785859204536, "value_loss": 3.6594990491867065}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:34:58.136868+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992098.13688, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9026651813280315, "elapsed_seconds": 6849.594652780914, "env_steps_per_second": 10511.570924969732, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21140724236978128, "eta_seconds": 761.0660725312126, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00014537982395014526, "chaser_grad_norm": 0.20405242964625359, "chaser_new_fields": 82.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.380000591278076, "chaser_subspace_norm": 102.37442779541016, "collisions_per_episode": 2.6500000953674316, "entropy": 0.04146633483469486, "explorer_grad_norm": 0.7894597426056862, "explorer_new_fields": 77.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 64.9749984741211, "explorer_subspace_norm": 36.01180648803711, "final_distance": 5.412384986877441, "max_param_abs": 1.3366508483886719, "policy_loss": -0.000498561928514718, "value_loss": 3.823637366294861}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:38:09.447719+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992289.4477334, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9546689239208355, "elapsed_seconds": 7036.808126115007, "env_steps_per_second": 10516.131557626411, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15848666950709475, "eta_seconds": 570.5520102255412, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 5.054187688694617e-05, "chaser_grad_norm": 0.06384878046810627, "chaser_new_fields": 81.875, "chaser_partner_vision": 0.0, "chaser_return": 6.220000743865967, "chaser_subspace_norm": 106.35965728759766, "collisions_per_episode": 2.75, "entropy": 0.04580217786133289, "explorer_grad_norm": 0.5481420829892159, "explorer_new_fields": 76.125, "explorer_partner_vision": 0.0, "explorer_return": 63.587501525878906, "explorer_subspace_norm": 37.49840545654297, "final_distance": 5.820192337036133, "max_param_abs": 1.35703444480896, "policy_loss": -0.0006632505802546262, "value_loss": 3.751565456390381}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:41:16.661191+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992476.661202, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0065912279688765, "elapsed_seconds": 7223.728420687956, "env_steps_per_second": 10520.882787113707, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10561006462994088, "eta_seconds": 380.19623266778717, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005322255128064057, "chaser_grad_norm": 0.2320895530283451, "chaser_new_fields": 81.75, "chaser_partner_vision": 0.0, "chaser_return": 6.225000858306885, "chaser_subspace_norm": 107.26871490478516, "collisions_per_episode": 2.450000047683716, "entropy": 0.05442480277270079, "explorer_grad_norm": 0.5851130858063698, "explorer_new_fields": 72.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 58.26250076293945, "explorer_subspace_norm": 36.695377349853516, "final_distance": 6.271575927734375, "max_param_abs": 1.3985481262207031, "policy_loss": -0.0008786491220789117, "value_loss": 3.7265177965164185}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:44:23.581486+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992663.5815008, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.059332836017776, "elapsed_seconds": 7413.598209663993, "env_steps_per_second": 10521.206814030349, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05280340605173785, "eta_seconds": 190.09226178625624, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0002307543342075502, "chaser_grad_norm": 0.0891418308019638, "chaser_new_fields": 82.75, "chaser_partner_vision": 0.0, "chaser_return": 6.385001182556152, "chaser_subspace_norm": 109.34814453125, "collisions_per_episode": 2.4750001430511475, "entropy": 0.057237508706748486, "explorer_grad_norm": 0.47530101239681244, "explorer_new_fields": 75.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 63.13750076293945, "explorer_subspace_norm": 34.646976470947266, "final_distance": 5.310695171356201, "max_param_abs": 1.4037516117095947, "policy_loss": -0.0008539053841971622, "value_loss": 3.6387414932250977}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:47:33.451276+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992853.4512906, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1043678205935996, "elapsed_seconds": 7575.7241541369585, "env_steps_per_second": 10560.046587270937, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:50:15.577218+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993015.577229, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1043706129752615, "elapsed_seconds": 7575.734206710942, "env_steps_per_second": 10560.032574681967, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:50:15.587272+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993015.5872822, "update": 20000} diff --git a/non_social/seed_0007/attempt_01/status.json b/non_social/seed_0007/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..6ba3c566ad5d8bbef013bf668aa83a641c02a851 --- /dev/null +++ b/non_social/seed_0007/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1043706129752615, + "elapsed_seconds": 7575.734206710942, + "env_steps_per_second": 10560.032574681967, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0005693047922270544, + "chaser_grad_norm": 0.08362068422138691, + "chaser_new_fields": 82.0250015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 6.3000006675720215, + "chaser_subspace_norm": 109.71160888671875, + "collisions_per_episode": 2.174999952316284, + "entropy": 0.05865261983126402, + "explorer_grad_norm": 0.6362885981798172, + "explorer_new_fields": 60.57500076293945, + "explorer_partner_vision": 0.0, + "explorer_return": 40.07500076293945, + "explorer_subspace_norm": 35.61513900756836, + "final_distance": 4.870059967041016, + "max_param_abs": 1.4202895164489746, + "policy_loss": -0.0008471746835780891, + "value_loss": 4.014994144439697 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:15.587272+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782993015.5872822, + "update": 20000 +} diff --git a/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b0a788f9b3245420bdee569da4f12d9513ed9e54 --- /dev/null +++ b/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d5ff79b8953d863166e3771c86afe125054881d88328fa27cd20f2770011ed51 +size 11460392 diff --git a/non_social/seed_0008/attempt_01/config.json b/non_social/seed_0008/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..46ac9d05b7bd5b4002c0b31b4dca84ef25970737 --- /dev/null +++ b/non_social/seed_0008/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", + "seed": 8, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0008/attempt_01/metrics.jsonl b/non_social/seed_0008/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..2567e5cac3ab7999919ff02b222e5690a022a58d --- /dev/null +++ b/non_social/seed_0008/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005259550897365747, "elapsed_seconds": 1.8934383230516687, "env_steps_per_second": 2112.5589100537322, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.518575839641755, "eta_seconds": 37866.87302271032, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 4.754632600975639e-05, "chaser_grad_norm": 0.4121139124035835, "chaser_new_fields": 33.525001525878906, "chaser_partner_vision": 0.0, "chaser_return": -3.325000524520874, "chaser_subspace_norm": 0.003984928131103516, "collisions_per_episode": 1.2000000476837158, "entropy": 2.772473454475403, "explorer_grad_norm": 0.15670402720570564, "explorer_new_fields": 33.150001525878906, "explorer_partner_vision": 0.0, "explorer_return": -0.4625000059604645, "explorer_subspace_norm": 0.0023247271310538054, "final_distance": 5.334477424621582, "max_param_abs": 0.15427866578102112, "policy_loss": -0.0006363048680952943, "value_loss": 3.079321026802063}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:44:01.529690+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985441.529705, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05249163369334484, "elapsed_seconds": 188.96988129604142, "env_steps_per_second": 10583.697181175592, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.047173714040449, "eta_seconds": 7369.825370545616, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00016925402738721118, "chaser_grad_norm": 0.14446954242885113, "chaser_new_fields": 73.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 4.650001049041748, "chaser_subspace_norm": 13.69598388671875, "collisions_per_episode": 2.8500001430511475, "entropy": 0.30062971264123917, "explorer_grad_norm": 0.23287267237901688, "explorer_new_fields": 72.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.32500076293945, "explorer_subspace_norm": 16.184181213378906, "final_distance": 5.430311679840088, "max_param_abs": 0.33060842752456665, "policy_loss": -0.002013030191418963, "value_loss": 4.428705334663391}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:47:08.606131+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985628.6061435, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10592686404444976, "elapsed_seconds": 381.3367105600191, "env_steps_per_second": 10489.41759141344, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0126104168445456, "eta_seconds": 7245.397500640364, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00032976457335287535, "chaser_grad_norm": 0.12632020562887192, "chaser_new_fields": 72.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 4.255000591278076, "chaser_subspace_norm": 24.848148345947266, "collisions_per_episode": 4.150000095367432, "entropy": 0.163105309009552, "explorer_grad_norm": 0.30370694771409035, "explorer_new_fields": 69.25, "explorer_partner_vision": 0.0, "explorer_return": 52.97500228881836, "explorer_subspace_norm": 21.950775146484375, "final_distance": 5.335358619689941, "max_param_abs": 0.35550814867019653, "policy_loss": -0.0015850234022423138, "value_loss": 4.264338970184326}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:50:20.972962+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985820.9729767, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15863393294359815, "elapsed_seconds": 571.0821585969534, "env_steps_per_second": 10506.369196931184, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.956485172971044, "eta_seconds": 7043.346622695758, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00011278843783912862, "chaser_grad_norm": 0.38048239797353745, "chaser_new_fields": 75.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.900001049041748, "chaser_subspace_norm": 34.497581481933594, "collisions_per_episode": 3.7250001430511475, "entropy": 0.12886571884155273, "explorer_grad_norm": 0.1848500780761242, "explorer_new_fields": 66.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 48.400001525878906, "explorer_subspace_norm": 23.64179229736328, "final_distance": 5.173815727233887, "max_param_abs": 0.36424916982650757, "policy_loss": -0.0010876346451702545, "value_loss": 3.9115403294563293}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:30.718409+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986010.7184222, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21073516378250334, "elapsed_seconds": 758.6465896170121, "env_steps_per_second": 10545.094526871391, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8966164740425302, "eta_seconds": 6827.819306553109, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00010625904829308919, "chaser_grad_norm": 0.31935570761561394, "chaser_new_fields": 74.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 4.835000514984131, "chaser_subspace_norm": 45.31084442138672, "collisions_per_episode": 3.424999952316284, "entropy": 0.09126113913953304, "explorer_grad_norm": 0.4457266628742218, "explorer_new_fields": 67.375, "explorer_partner_vision": 0.0, "explorer_return": 50.35000228881836, "explorer_subspace_norm": 27.325475692749023, "final_distance": 5.649575233459473, "max_param_abs": 0.4153917133808136, "policy_loss": -0.0005672876851221531, "value_loss": 4.1097763776779175}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:38.282840+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986198.2828538, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2642951669180507, "elapsed_seconds": 951.4626009049825, "env_steps_per_second": 10510.13459750127, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8500661684263549, "eta_seconds": 6660.238206334878, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0009062783354628068, "chaser_grad_norm": 0.2573314495384693, "chaser_new_fields": 77.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.355000972747803, "chaser_subspace_norm": 50.96757888793945, "collisions_per_episode": 3.125, "entropy": 0.09703690558671951, "explorer_grad_norm": 0.48943834751844406, "explorer_new_fields": 67.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 50.53750228881836, "explorer_subspace_norm": 28.96225929260254, "final_distance": 4.776935577392578, "max_param_abs": 0.44517865777015686, "policy_loss": -0.001581848564790267, "value_loss": 3.742652118206024}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:51.098852+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986391.0988653, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3172790927558284, "elapsed_seconds": 1142.204733920982, "env_steps_per_second": 10505.997430781234, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.797914858949694, "eta_seconds": 6472.493492218899, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 2.621762969978736e-05, "chaser_grad_norm": 0.23211281187832355, "chaser_new_fields": 78.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.515000820159912, "chaser_subspace_norm": 55.347816467285156, "collisions_per_episode": 2.9750001430511475, "entropy": 0.06468199379742146, "explorer_grad_norm": 0.24866489693522453, "explorer_new_fields": 67.125, "explorer_partner_vision": 0.0, "explorer_return": 49.712501525878906, "explorer_subspace_norm": 31.527503967285156, "final_distance": 4.6293230056762695, "max_param_abs": 0.5851640701293945, "policy_loss": -0.0006789298250229336, "value_loss": 3.708665192127228}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:03:01.840984+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986581.8409986, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.37014522816888834, "elapsed_seconds": 1332.522821407998, "env_steps_per_second": 10506.386663762372, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7449703613676166, "eta_seconds": 6281.89330092342, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00036730787367353357, "chaser_grad_norm": 0.21600015833973885, "chaser_new_fields": 76.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.250000953674316, "chaser_subspace_norm": 61.89194869995117, "collisions_per_episode": 2.9000000953674316, "entropy": 0.07542024180293083, "explorer_grad_norm": 0.8061941638588905, "explorer_new_fields": 66.875, "explorer_partner_vision": 0.0, "explorer_return": 49.525001525878906, "explorer_subspace_norm": 34.09337615966797, "final_distance": 4.518491268157959, "max_param_abs": 0.630811870098114, "policy_loss": -0.0007589059532620013, "value_loss": 3.540228247642517}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:06:12.159070+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986772.1590834, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4218972936016831, "elapsed_seconds": 1518.830256966059, "env_steps_per_second": 10534.422741854522, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6875891744067324, "eta_seconds": 6075.321027864236, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0003093269264735632, "chaser_grad_norm": 0.05942075606435537, "chaser_new_fields": 78.375, "chaser_partner_vision": 0.0, "chaser_return": 5.5400004386901855, "chaser_subspace_norm": 67.25521087646484, "collisions_per_episode": 2.875, "entropy": 0.07334532588720322, "explorer_grad_norm": 0.3758119270205498, "explorer_new_fields": 67.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 50.57500076293945, "explorer_subspace_norm": 33.30562210083008, "final_distance": 5.884316444396973, "max_param_abs": 0.6320380568504333, "policy_loss": -0.0008024419229499813, "value_loss": 3.7398738861083984}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:09:18.466507+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986958.4665222, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4741110587077694, "elapsed_seconds": 1706.7998113479698, "env_steps_per_second": 10546.052255410223, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6330492022156502, "eta_seconds": 5878.977127976341, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0006056344681750758, "chaser_grad_norm": 0.15694402903318405, "chaser_new_fields": 78.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.630001068115234, "chaser_subspace_norm": 73.3217544555664, "collisions_per_episode": 2.450000047683716, "entropy": 0.13684245198965073, "explorer_grad_norm": 0.7262015156447887, "explorer_new_fields": 68.25, "explorer_partner_vision": 0.0, "explorer_return": 51.587501525878906, "explorer_subspace_norm": 30.86418914794922, "final_distance": 4.977447509765625, "max_param_abs": 0.6942281723022461, "policy_loss": -0.0012669981275763575, "value_loss": 3.5673720240592957}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:26.436064+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987146.436078, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5277853027694316, "elapsed_seconds": 1900.0270899699535, "env_steps_per_second": 10526.165708677487, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5833559083082946, "eta_seconds": 5700.081269909861, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0004787933788036325, "chaser_grad_norm": 0.12030955404043198, "chaser_new_fields": 78.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.560000896453857, "chaser_subspace_norm": 74.74514770507812, "collisions_per_episode": 2.5, "entropy": 0.12300128862261772, "explorer_grad_norm": 0.6194204278290272, "explorer_new_fields": 68.875, "explorer_partner_vision": 0.0, "explorer_return": 52.29999923706055, "explorer_subspace_norm": 30.457660675048828, "final_distance": 5.5447306632995605, "max_param_abs": 0.6979689002037048, "policy_loss": -0.00038201219207278925, "value_loss": 3.616903245449066}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:39.663339+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987339.6633525, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5797138622980694, "elapsed_seconds": 2086.96990427305, "env_steps_per_second": 10541.599069040345, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5283365460585467, "eta_seconds": 5502.011565810768, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": -7.408035091860388e-05, "chaser_grad_norm": 0.281186006963253, "chaser_new_fields": 82.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 75.3775405883789, "collisions_per_episode": 2.174999952316284, "entropy": 0.12395346164703369, "explorer_grad_norm": 1.110318273305893, "explorer_new_fields": 67.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 50.35000228881836, "explorer_subspace_norm": 33.52949142456055, "final_distance": 4.978503227233887, "max_param_abs": 0.6979689002037048, "policy_loss": -0.00022923882305603271, "value_loss": 3.6345176696777344}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:18:46.606153+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987526.6061652, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6323410607099909, "elapsed_seconds": 2276.427818555967, "env_steps_per_second": 10542.833734664251, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4754624749899787, "eta_seconds": 5311.6649099639235, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006116828866263441, "chaser_grad_norm": 0.09661844093352556, "chaser_new_fields": 83.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.54500150680542, "chaser_subspace_norm": 86.88473510742188, "collisions_per_episode": 1.600000023841858, "entropy": 0.12295115552842617, "explorer_grad_norm": 0.42696069926023483, "explorer_new_fields": 65.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 48.025001525878906, "explorer_subspace_norm": 33.49871826171875, "final_distance": 5.205170631408691, "max_param_abs": 0.698940634727478, "policy_loss": -0.00023481844546324737, "value_loss": 3.479794979095459}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:21:56.064070+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987716.0640838, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6852637503711028, "elapsed_seconds": 2466.94950133597, "env_steps_per_second": 10539.332072229192, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4232400969245982, "eta_seconds": 5123.664348928553, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003974755190575621, "chaser_grad_norm": 0.14166885055601597, "chaser_new_fields": 83.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.645000457763672, "chaser_subspace_norm": 84.30269622802734, "collisions_per_episode": 1.3250000476837158, "entropy": 0.11714527010917664, "explorer_grad_norm": 0.7173113003373146, "explorer_new_fields": 68.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.63750076293945, "explorer_subspace_norm": 34.35336685180664, "final_distance": 5.8533124923706055, "max_param_abs": 0.7168130278587341, "policy_loss": -0.0014581020986597082, "value_loss": 3.7268956899642944}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:25:06.585750+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987906.585762, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7370353669483383, "elapsed_seconds": 2653.327321014018, "env_steps_per_second": 10552.787731179464, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.368779967189771, "eta_seconds": 4927.607881883176, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0003065439267603276, "chaser_grad_norm": 0.10413557942956686, "chaser_new_fields": 82.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.355000972747803, "chaser_subspace_norm": 87.8616714477539, "collisions_per_episode": 2.2750000953674316, "entropy": 0.11093438975512981, "explorer_grad_norm": 0.6659631952643394, "explorer_new_fields": 69.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 53.6875, "explorer_subspace_norm": 34.531986236572266, "final_distance": 4.769343852996826, "max_param_abs": 0.7794198989868164, "policy_loss": -0.000987638846504879, "value_loss": 3.5082451701164246}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:28:12.963572+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988092.9635887, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.788929970683627, "elapsed_seconds": 2840.147894461057, "env_steps_per_second": 10562.830216872479, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3148832844727116, "eta_seconds": 4733.579824101762, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0002495239944116656, "chaser_grad_norm": 0.07598771527409554, "chaser_new_fields": 83.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.615001201629639, "chaser_subspace_norm": 91.1243896484375, "collisions_per_episode": 1.850000023841858, "entropy": 0.12336103059351444, "explorer_grad_norm": 2.268488824367523, "explorer_new_fields": 68.75, "explorer_partner_vision": 0.0, "explorer_return": 52.150001525878906, "explorer_subspace_norm": 38.40073776245117, "final_distance": 4.4665422439575195, "max_param_abs": 0.8008120656013489, "policy_loss": -0.0003160947775810996, "value_loss": 3.3141525387763977}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:31:19.784143+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988279.7841547, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8406168637777834, "elapsed_seconds": 3026.22070960002, "env_steps_per_second": 10574.24526191597, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.260925295666675, "eta_seconds": 4539.33106440003, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00016047299792348868, "chaser_grad_norm": 0.1372926477342844, "chaser_new_fields": 82.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.330000400543213, "chaser_subspace_norm": 89.40558624267578, "collisions_per_episode": 2.0250000953674316, "entropy": 0.10477716289460659, "explorer_grad_norm": 2.721346437931061, "explorer_new_fields": 69.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 53.91250228881836, "explorer_subspace_norm": 40.943321228027344, "final_distance": 5.055099010467529, "max_param_abs": 0.8200756907463074, "policy_loss": -0.00047149774734833727, "value_loss": 3.4087679982185364}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:34:25.856960+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988465.8569741, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8937212799366615, "elapsed_seconds": 3217.3966077719815, "env_steps_per_second": 10567.5501484241, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2091523199143068, "eta_seconds": 4352.948351691504, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005901245169914054, "chaser_grad_norm": 0.09996180422604084, "chaser_new_fields": 83.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.645001411437988, "chaser_subspace_norm": 97.29917907714844, "collisions_per_episode": 2.200000047683716, "entropy": 0.09270570240914822, "explorer_grad_norm": 0.7610118463635445, "explorer_new_fields": 68.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 52.60000228881836, "explorer_subspace_norm": 40.31224060058594, "final_distance": 5.0274810791015625, "max_param_abs": 0.8080316185951233, "policy_loss": -0.001295726840149669, "value_loss": 3.3391451239585876}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:37:37.032858+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988657.0328708, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9467771115455414, "elapsed_seconds": 3408.397601563949, "env_steps_per_second": 10562.148026240055, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1571720252223283, "eta_seconds": 4165.819290800382, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0002142162214308213, "chaser_grad_norm": 0.045544703491032124, "chaser_new_fields": 82.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.385000705718994, "chaser_subspace_norm": 94.84537506103516, "collisions_per_episode": 1.9250000715255737, "entropy": 0.11676365695893764, "explorer_grad_norm": 1.1080102697014809, "explorer_new_fields": 68.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 52.375, "explorer_subspace_norm": 44.32368469238281, "final_distance": 5.008363246917725, "max_param_abs": 0.8217601180076599, "policy_loss": -0.0009213356246480231, "value_loss": 3.4127944111824036}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:40:48.033851+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988848.0338638, "update": 9000} +{"device": "cuda", "elapsed_hours": 0.9997739965880627, "elapsed_seconds": 3599.1863877170254, "env_steps_per_second": 10557.941686399718, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.105013364649964, "eta_seconds": 3978.0481127398702, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0006356161760887047, "chaser_grad_norm": 0.07004822045564651, "chaser_new_fields": 80.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.090001106262207, "chaser_subspace_norm": 94.5180435180664, "collisions_per_episode": 2.2750000953674316, "entropy": 0.08138025924563408, "explorer_grad_norm": 1.7572572827339172, "explorer_new_fields": 67.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 50.35000228881836, "explorer_subspace_norm": 47.94840621948242, "final_distance": 5.016236782073975, "max_param_abs": 0.8415794372558594, "policy_loss": -0.0009667087035483135, "value_loss": 3.550901770591736}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:43:58.822636+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989038.8226483, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0517289863627715, "elapsed_seconds": 3786.224350905977, "env_steps_per_second": 10564.614320973531, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0517289863627715, "eta_seconds": 3786.224350905977, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00010283939993138702, "chaser_grad_norm": 0.1543809473514557, "chaser_new_fields": 82.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.295001029968262, "chaser_subspace_norm": 99.1401138305664, "collisions_per_episode": 2.325000047683716, "entropy": 0.1387317068874836, "explorer_grad_norm": 0.4958537593483925, "explorer_new_fields": 67.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 50.23749923706055, "explorer_subspace_norm": 50.27497863769531, "final_distance": 5.395737648010254, "max_param_abs": 1.024712324142456, "policy_loss": -0.001239839154884237, "value_loss": 3.602066218852997}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:47:05.860601+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989225.860614, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.105616985437777, "elapsed_seconds": 3980.2211475759977, "env_steps_per_second": 10552.177490333295, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0003201296817983, "eta_seconds": 3601.152466854474, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -0.00014003355478196464, "chaser_grad_norm": 0.07403950113803148, "chaser_new_fields": 84.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7350006103515625, "chaser_subspace_norm": 101.34937286376953, "collisions_per_episode": 1.4250000715255737, "entropy": 0.08680791780352592, "explorer_grad_norm": 2.2947110533714294, "explorer_new_fields": 68.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 52.67499923706055, "explorer_subspace_norm": 50.815696716308594, "final_distance": 6.2726359367370605, "max_param_abs": 1.0405265092849731, "policy_loss": -0.0005693615898483273, "value_loss": 3.465144693851471}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:19.857396+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989419.8574078, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1582325217780696, "elapsed_seconds": 4169.6370784010505, "env_steps_per_second": 10552.477151530147, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9476447905456933, "eta_seconds": 3411.521245964496, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0007534046398861849, "chaser_grad_norm": 0.032103750854730606, "chaser_new_fields": 83.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.49500036239624, "chaser_subspace_norm": 106.50418090820312, "collisions_per_episode": 2.0250000953674316, "entropy": 0.10554543510079384, "explorer_grad_norm": 0.6380412876605988, "explorer_new_fields": 69.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 53.53750228881836, "explorer_subspace_norm": 52.28444290161133, "final_distance": 5.147563934326172, "max_param_abs": 1.2655736207962036, "policy_loss": -0.0015591749700263335, "value_loss": 3.2837368845939636}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:53:29.273329+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989609.2733433, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.211334609787219, "elapsed_seconds": 4360.804595233989, "env_steps_per_second": 10548.512091157289, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8953342767992489, "eta_seconds": 3223.2033964772963, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00022638231255889263, "chaser_grad_norm": 0.06078511942178011, "chaser_new_fields": 83.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.590001106262207, "chaser_subspace_norm": 104.25740814208984, "collisions_per_episode": 1.875, "entropy": 0.08638930134475231, "explorer_grad_norm": 1.8808839321136475, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 49.86249923706055, "explorer_subspace_norm": 51.196327209472656, "final_distance": 5.4833760261535645, "max_param_abs": 1.3564914464950562, "policy_loss": -7.382393271804233e-05, "value_loss": 3.4229121804237366}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:56:40.440845+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989800.4408581, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2636336978844418, "elapsed_seconds": 4549.081312383991, "env_steps_per_second": 10551.581012484723, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8424224652562945, "eta_seconds": 3032.72087492266, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00023287999600718479, "chaser_grad_norm": 0.13514600321650505, "chaser_new_fields": 84.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.765000820159912, "chaser_subspace_norm": 106.16870880126953, "collisions_per_episode": 2.2250001430511475, "entropy": 0.10444030351936817, "explorer_grad_norm": 0.6184110529720783, "explorer_new_fields": 69.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 53.650001525878906, "explorer_subspace_norm": 54.39659881591797, "final_distance": 5.379390239715576, "max_param_abs": 1.3571827411651611, "policy_loss": -0.00047393275885987407, "value_loss": 3.443018853664398}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:59:48.717562+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989988.717577, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.316743994015851, "elapsed_seconds": 4740.278378457064, "env_steps_per_second": 10547.903732243409, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7900463964095107, "eta_seconds": 2844.1670270742384, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0009416277081147409, "chaser_grad_norm": 0.12493641674518585, "chaser_new_fields": 82.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.42000150680542, "chaser_subspace_norm": 102.50790405273438, "collisions_per_episode": 1.399999976158142, "entropy": 0.08975577168166637, "explorer_grad_norm": 0.5014815032482147, "explorer_new_fields": 70.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 54.66250228881836, "explorer_subspace_norm": 51.16746139526367, "final_distance": 5.47189474105835, "max_param_abs": 1.3571856021881104, "policy_loss": -0.0007117555135636167, "value_loss": 3.40231055021286}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:02:59.914627+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990179.9146378, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3703398133472202, "elapsed_seconds": 4933.223328049993, "env_steps_per_second": 10540.775582636066, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7378752841100417, "eta_seconds": 2656.35102279615, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00030545931165778736, "chaser_grad_norm": 0.13994945120066404, "chaser_new_fields": 81.5, "chaser_partner_vision": 0.0, "chaser_return": 6.1800007820129395, "chaser_subspace_norm": 109.9525146484375, "collisions_per_episode": 2.3500001430511475, "entropy": 0.11504868231713772, "explorer_grad_norm": 1.3664233908057213, "explorer_new_fields": 69.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 52.97500228881836, "explorer_subspace_norm": 53.89692306518555, "final_distance": 5.133326053619385, "max_param_abs": 1.3931372165679932, "policy_loss": -0.0008943903929434782, "value_loss": 3.6017600893974304}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:06:12.859580+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990372.8595958, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4236447368130514, "elapsed_seconds": 5125.121052526985, "env_steps_per_second": 10536.336497530108, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6854585769840618, "eta_seconds": 2467.6508771426224, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00019815828474852026, "chaser_grad_norm": 0.06455073785036802, "chaser_new_fields": 85.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.915000915527344, "chaser_subspace_norm": 116.67318725585938, "collisions_per_episode": 1.899999976158142, "entropy": 0.11709864996373653, "explorer_grad_norm": 1.3403055667877197, "explorer_new_fields": 70.875, "explorer_partner_vision": 0.0, "explorer_return": 55.63750076293945, "explorer_subspace_norm": 50.90962600708008, "final_distance": 5.89241886138916, "max_param_abs": 1.3904222249984741, "policy_loss": -0.0006219206412083622, "value_loss": 3.192169725894928}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:09:24.757302+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990564.7573152, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.476959545706664, "elapsed_seconds": 5317.05436454399, "env_steps_per_second": 10532.147343353854, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6329826624457131, "eta_seconds": 2278.737584804567, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 1.5411911990628746e-05, "chaser_grad_norm": 0.0640657152980566, "chaser_new_fields": 81.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.14500093460083, "chaser_subspace_norm": 113.47754669189453, "collisions_per_episode": 2.6500000953674316, "entropy": 0.12624214962124825, "explorer_grad_norm": 1.1085692644119263, "explorer_new_fields": 70.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 54.4375, "explorer_subspace_norm": 51.47956848144531, "final_distance": 4.886295318603516, "max_param_abs": 1.395499348640442, "policy_loss": -0.0005137781260984298, "value_loss": 3.364220976829529}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:12:36.690616+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990756.6906307, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5296791687133373, "elapsed_seconds": 5506.845007368014, "env_steps_per_second": 10532.346547323834, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5802231329602314, "eta_seconds": 2088.8032786568333, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0005118708604719413, "chaser_grad_norm": 0.048516035079956055, "chaser_new_fields": 81.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.320001125335693, "chaser_subspace_norm": 115.04212951660156, "collisions_per_episode": 1.6750000715255737, "entropy": 0.09706471674144268, "explorer_grad_norm": 0.5769659578800201, "explorer_new_fields": 70.0, "explorer_partner_vision": 0.0, "explorer_return": 54.36249923706055, "explorer_subspace_norm": 55.18115997314453, "final_distance": 5.316706657409668, "max_param_abs": 1.4108127355575562, "policy_loss": -0.0004109378027479238, "value_loss": 3.2147222757339478}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:15:46.481256+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990946.4812675, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5823092335941813, "elapsed_seconds": 5696.313240939053, "env_steps_per_second": 10533.128615326787, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5274364111980604, "eta_seconds": 1898.7710803130174, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0003081051429691395, "chaser_grad_norm": 0.06997780967503786, "chaser_new_fields": 76.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.250000953674316, "chaser_subspace_norm": 116.75130462646484, "collisions_per_episode": 2.8500001430511475, "entropy": 0.11324145458638668, "explorer_grad_norm": 3.539816439151764, "explorer_new_fields": 71.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 56.650001525878906, "explorer_subspace_norm": 57.0086555480957, "final_distance": 5.537540435791016, "max_param_abs": 1.4194040298461914, "policy_loss": -0.00035141751862965975, "value_loss": 3.4655900597572327}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:18:55.949490+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991135.9495046, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6340214860494597, "elapsed_seconds": 5882.477349778055, "env_steps_per_second": 10539.77708938212, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4743933346595206, "eta_seconds": 1707.8160047742742, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00047327737344553256, "chaser_grad_norm": 0.06632671691477299, "chaser_new_fields": 76.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 5.085000514984131, "chaser_subspace_norm": 120.0212173461914, "collisions_per_episode": 2.625, "entropy": 0.10908017121255398, "explorer_grad_norm": 1.2005017846822739, "explorer_new_fields": 70.75, "explorer_partner_vision": 0.0, "explorer_return": 55.150001525878906, "explorer_subspace_norm": 57.642024993896484, "final_distance": 5.289363861083984, "max_param_abs": 1.4299330711364746, "policy_loss": -0.0005172209861257215, "value_loss": 3.261428475379944}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:22:02.113598+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991322.1136096, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.687708684888332, "elapsed_seconds": 6075.751265597995, "env_steps_per_second": 10533.676775476244, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.421927171222083, "eta_seconds": 1518.9378163994988, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": -0.00012615988844547066, "chaser_grad_norm": 0.04006043076515198, "chaser_new_fields": 76.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.075000762939453, "chaser_subspace_norm": 119.19915771484375, "collisions_per_episode": 2.875, "entropy": 0.12961988151073456, "explorer_grad_norm": 1.3424361869692802, "explorer_new_fields": 70.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 54.47500228881836, "explorer_subspace_norm": 55.98262023925781, "final_distance": 4.544209957122803, "max_param_abs": 1.429936170578003, "policy_loss": -0.0009435678138487447, "value_loss": 3.4368958473205566}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:25:15.387515+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991515.3875294, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.741354218290847, "elapsed_seconds": 6268.87518584705, "env_steps_per_second": 10528.204509320134, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3693781675162403, "eta_seconds": 1329.761403058465, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00017138751045564504, "chaser_grad_norm": 0.08112307731062174, "chaser_new_fields": 76.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.1750006675720215, "chaser_subspace_norm": 122.82561492919922, "collisions_per_episode": 2.7750000953674316, "entropy": 0.12553239054977894, "explorer_grad_norm": 1.5775202512741089, "explorer_new_fields": 72.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 57.400001525878906, "explorer_subspace_norm": 55.62919235229492, "final_distance": 5.4820709228515625, "max_param_abs": 1.429936170578003, "policy_loss": -0.0004014897489135194, "value_loss": 3.3123549222946167}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:28:28.511436+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991708.511449, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7953756610611, "elapsed_seconds": 6463.35237981996, "env_steps_per_second": 10520.856051777602, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3168309990107824, "eta_seconds": 1140.5915964388166, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0006751127057105025, "chaser_grad_norm": 0.037811532616615295, "chaser_new_fields": 77.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.315001010894775, "chaser_subspace_norm": 120.45658111572266, "collisions_per_episode": 3.075000047683716, "entropy": 0.08956804871559143, "explorer_grad_norm": 1.2036668956279755, "explorer_new_fields": 71.5, "explorer_partner_vision": 0.0, "explorer_return": 56.35000228881836, "explorer_subspace_norm": 61.887107849121094, "final_distance": 4.991678714752197, "max_param_abs": 1.4014315605163574, "policy_loss": -0.0011333116936640408, "value_loss": 3.466517388820648}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:31:42.988631+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991902.988647, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8479744270874976, "elapsed_seconds": 6652.7079375149915, "env_steps_per_second": 10522.031127394921, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26399634672678535, "eta_seconds": 950.3868482164273, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0004689219763659991, "chaser_grad_norm": 0.07354088593274355, "chaser_new_fields": 74.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 4.695000648498535, "chaser_subspace_norm": 118.07134246826172, "collisions_per_episode": 2.875, "entropy": 0.0901863370090723, "explorer_grad_norm": 1.2921668142080307, "explorer_new_fields": 72.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 58.337501525878906, "explorer_subspace_norm": 61.288455963134766, "final_distance": 5.3530426025390625, "max_param_abs": 1.42307448387146, "policy_loss": -0.00069365562527679, "value_loss": 3.3632800579071045}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:34:52.344189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992092.3442035, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.899906994776393, "elapsed_seconds": 6839.665181195014, "env_steps_per_second": 10526.831079093887, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.211100777197377, "eta_seconds": 759.9627979105571, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0007138783019062789, "chaser_grad_norm": 0.11010011099278927, "chaser_new_fields": 75.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.890000820159912, "chaser_subspace_norm": 130.8722381591797, "collisions_per_episode": 2.950000047683716, "entropy": 0.09489231742918491, "explorer_grad_norm": 0.9460238292813301, "explorer_new_fields": 71.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 57.17499923706055, "explorer_subspace_norm": 60.435272216796875, "final_distance": 4.605357646942139, "max_param_abs": 1.4230958223342896, "policy_loss": -0.000962960884450581, "value_loss": 3.4205397367477417}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:37:59.301431+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992279.3014462, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9534467903658193, "elapsed_seconds": 7032.408445316949, "env_steps_per_second": 10522.710757689052, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1583875775972286, "eta_seconds": 570.1952793500229, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0007010686587220344, "chaser_grad_norm": 0.10755470767617226, "chaser_new_fields": 75.25, "chaser_partner_vision": 0.0, "chaser_return": 4.960000514984131, "chaser_subspace_norm": 128.18035888671875, "collisions_per_episode": 2.375, "entropy": 0.07398054748773575, "explorer_grad_norm": 1.6895762979984283, "explorer_new_fields": 71.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 57.400001525878906, "explorer_subspace_norm": 63.78392791748047, "final_distance": 4.197110652923584, "max_param_abs": 1.416873812675476, "policy_loss": -0.0009155682309085478, "value_loss": 3.388755738735199}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:41:12.044694+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992472.0447066, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0072649444869604, "elapsed_seconds": 7226.153800153057, "env_steps_per_second": 10517.3515679102, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10564552339405053, "eta_seconds": 380.3238842185819, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00031048283401413634, "chaser_grad_norm": 0.15179618634283543, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.005000591278076, "chaser_subspace_norm": 144.87802124023438, "collisions_per_episode": 2.049999952316284, "entropy": 0.09089219942688942, "explorer_grad_norm": 2.5168567299842834, "explorer_new_fields": 70.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 55.60000228881836, "explorer_subspace_norm": 66.1735610961914, "final_distance": 5.284684181213379, "max_param_abs": 1.4185112714767456, "policy_loss": -0.0009407151557994098, "value_loss": 3.722986102104187}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:44:25.790051+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992665.7900658, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.060189794643601, "elapsed_seconds": 7416.683260716964, "env_steps_per_second": 10516.830402227506, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05282537934983592, "eta_seconds": 190.17136565940933, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00017237227471436634, "chaser_grad_norm": 0.07433113176375628, "chaser_new_fields": 80.125, "chaser_partner_vision": 0.0, "chaser_return": 5.865001201629639, "chaser_subspace_norm": 140.71307373046875, "collisions_per_episode": 2.75, "entropy": 0.07999558933079243, "explorer_grad_norm": 1.3726843819022179, "explorer_new_fields": 71.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 56.57500076293945, "explorer_subspace_norm": 63.62240219116211, "final_distance": 4.568384647369385, "max_param_abs": 1.4185677766799927, "policy_loss": -0.0003749398289372152, "value_loss": 3.6020233035087585}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:47:36.319509+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992856.319521, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.10492606138361, "elapsed_seconds": 7577.733820980997, "env_steps_per_second": 10557.245990681073, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:50:17.370069+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993017.3700802, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.10492892708251, "elapsed_seconds": 7577.744137497037, "env_steps_per_second": 10557.231617802336, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:50:17.380387+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993017.380397, "update": 20000} diff --git a/non_social/seed_0008/attempt_01/status.json b/non_social/seed_0008/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..2368f786446327580c1594ff90c975b3afe8c158 --- /dev/null +++ b/non_social/seed_0008/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.10492892708251, + "elapsed_seconds": 7577.744137497037, + "env_steps_per_second": 10557.231617802336, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 3.927184380664306e-06, + "chaser_grad_norm": 0.07202051393687725, + "chaser_new_fields": 81.3499984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.145000457763672, + "chaser_subspace_norm": 132.7053680419922, + "collisions_per_episode": 1.975000023841858, + "entropy": 0.13418130949139595, + "explorer_grad_norm": 0.6797178164124489, + "explorer_new_fields": 68.7249984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 52.48749923706055, + "explorer_subspace_norm": 61.86991500854492, + "final_distance": 4.521096229553223, + "max_param_abs": 1.4186153411865234, + "policy_loss": -0.0009259324722812323, + "value_loss": 4.002953112125397 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:17.380387+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782993017.380397, + "update": 20000 +} diff --git a/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors b/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..49b9e21c404d5fa935fc1f5ea0dcd26453e047b4 --- /dev/null +++ b/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:def8dbb315937927f72b78afdd9894dd5613ef88bc1bdbdb7b654ba5e17e690d +size 11460400 diff --git a/non_social/seed_0009/attempt_01/config.json b/non_social/seed_0009/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..7ccf3b0e88e1c2d396a34cd8002980715cc677c7 --- /dev/null +++ b/non_social/seed_0009/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "none", + "task": "non_social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", + "seed": 9, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/non_social/seed_0009/attempt_01/metrics.jsonl b/non_social/seed_0009/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..dfd5b33b51b85e114865379a65b716e17bfdd6e3 --- /dev/null +++ b/non_social/seed_0009/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005488903238761446, "elapsed_seconds": 1.9760051659541205, "env_steps_per_second": 2024.286205784582, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.977257587199015, "eta_seconds": 39518.127313916455, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 8.324658641868654e-05, "chaser_grad_norm": 0.3642922267317772, "chaser_new_fields": 34.42499923706055, "chaser_partner_vision": 0.0, "chaser_return": -3.135000467300415, "chaser_subspace_norm": 0.002859547035768628, "collisions_per_episode": 1.350000023841858, "entropy": 2.772511601448059, "explorer_grad_norm": 0.10375835932791233, "explorer_new_fields": 35.650001525878906, "explorer_partner_vision": 0.0, "explorer_return": 3.1000001430511475, "explorer_subspace_norm": 0.002681385725736618, "final_distance": 4.754347324371338, "max_param_abs": 0.15429173409938812, "policy_loss": -0.0005586856793975414, "value_loss": 2.9997570514678955}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:44:01.717349+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985441.7173617, "update": 1} +{"device": "cuda", "elapsed_hours": 0.054379662316674106, "elapsed_seconds": 195.7667843400268, "env_steps_per_second": 10216.237686809043, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1208068303502903, "eta_seconds": 7634.904589261045, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0007525032333211001, "chaser_grad_norm": 0.1295870915055275, "chaser_new_fields": 72.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.380000591278076, "chaser_subspace_norm": 13.003899574279785, "collisions_per_episode": 3.049999952316284, "entropy": 0.4666449874639511, "explorer_grad_norm": 0.47834302484989166, "explorer_new_fields": 63.97500228881836, "explorer_partner_vision": 0.0, "explorer_return": 45.28750228881836, "explorer_subspace_norm": 8.475234985351562, "final_distance": 5.086209297180176, "max_param_abs": 0.31064581871032715, "policy_loss": -0.0020309365144939306, "value_loss": 4.75316047668457}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:47:15.508130+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985635.5081434, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10749254430582773, "elapsed_seconds": 386.9731595009798, "env_steps_per_second": 10336.634212972778, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.042358341810727, "eta_seconds": 7352.490030518617, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00012754829878769414, "chaser_grad_norm": 0.2244296856224537, "chaser_new_fields": 76.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.1800007820129395, "chaser_subspace_norm": 17.93508529663086, "collisions_per_episode": 3.450000047683716, "entropy": 0.3611351102590561, "explorer_grad_norm": 0.3530866950750351, "explorer_new_fields": 66.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 48.400001525878906, "explorer_subspace_norm": 8.996316909790039, "final_distance": 4.814425945281982, "max_param_abs": 0.3237481117248535, "policy_loss": -0.0014594234691500674, "value_loss": 4.740478277206421}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:50:26.714503+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985826.7145166, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16100657242638086, "elapsed_seconds": 579.623660734971, "env_steps_per_second": 10351.544297539398, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9857477265920305, "eta_seconds": 7148.691815731309, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 5.388649399584722e-06, "chaser_grad_norm": 0.2633400224149227, "chaser_new_fields": 74.625, "chaser_partner_vision": 0.0, "chaser_return": 4.785000801086426, "chaser_subspace_norm": 20.397611618041992, "collisions_per_episode": 2.575000047683716, "entropy": 0.3132750168442726, "explorer_grad_norm": 0.6032910346984863, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.1875, "explorer_subspace_norm": 10.505915641784668, "final_distance": 6.120945453643799, "max_param_abs": 0.3252468705177307, "policy_loss": -0.0015240332195549655, "value_loss": 4.562912583351135}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:39.365002+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986019.3650138, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21344087832387434, "elapsed_seconds": 768.3871619659476, "env_steps_per_second": 10411.418092321712, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.920967904914869, "eta_seconds": 6915.484457693528, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0003194324828947881, "chaser_grad_norm": 0.31822415068745613, "chaser_new_fields": 76.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.240001201629639, "chaser_subspace_norm": 29.54732894897461, "collisions_per_episode": 3.049999952316284, "entropy": 0.23961805924773216, "explorer_grad_norm": 0.9347753897309303, "explorer_new_fields": 69.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 53.5, "explorer_subspace_norm": 11.83540153503418, "final_distance": 4.50183629989624, "max_param_abs": 0.36849361658096313, "policy_loss": -0.0013278836022578844, "value_loss": 4.314780354499817}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:48.128506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986208.1285188, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2650195285063819, "elapsed_seconds": 954.0703026229749, "env_steps_per_second": 10481.40789259191, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8551366995446734, "eta_seconds": 6678.492118360824, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00023550841499427738, "chaser_grad_norm": 0.2011011205613613, "chaser_new_fields": 78.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.470000743865967, "chaser_subspace_norm": 41.14895248413086, "collisions_per_episode": 2.924999952316284, "entropy": 0.21957940235733986, "explorer_grad_norm": 0.3941175043582916, "explorer_new_fields": 66.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 49.0, "explorer_subspace_norm": 12.696320533752441, "final_distance": 5.125357627868652, "max_param_abs": 0.4827040731906891, "policy_loss": -0.001310012520264081, "value_loss": 4.246262907981873}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:53.811643+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986393.8116558, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.317056712508622, "elapsed_seconds": 1141.4041650310392, "env_steps_per_second": 10513.366227004852, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7966547042155248, "eta_seconds": 6467.956935175889, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 5.0292346280178535e-05, "chaser_grad_norm": 0.28287162259221077, "chaser_new_fields": 82.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.440000057220459, "chaser_subspace_norm": 45.37813186645508, "collisions_per_episode": 2.299999952316284, "entropy": 0.1720152273774147, "explorer_grad_norm": 0.6429248973727226, "explorer_new_fields": 70.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 54.73749923706055, "explorer_subspace_norm": 14.35446548461914, "final_distance": 5.009875774383545, "max_param_abs": 0.535347044467926, "policy_loss": -0.0014211593733506334, "value_loss": 3.8803778290748596}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:03:01.145507+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986581.1455219, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3691531043750001, "elapsed_seconds": 1328.9511757500004, "env_steps_per_second": 10534.623284485247, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7402932063392862, "eta_seconds": 6265.05554282143, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 4.60841988100813e-05, "chaser_grad_norm": 0.11920395959168673, "chaser_new_fields": 77.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.340000629425049, "chaser_subspace_norm": 48.65547561645508, "collisions_per_episode": 2.575000047683716, "entropy": 0.15424678474664688, "explorer_grad_norm": 1.22721566259861, "explorer_new_fields": 71.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 56.95000076293945, "explorer_subspace_norm": 14.79220199584961, "final_distance": 5.212035179138184, "max_param_abs": 0.5754305124282837, "policy_loss": -0.0008111786241684626, "value_loss": 4.039854526519775}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:06:08.692517+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986768.6925268, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4217239715455475, "elapsed_seconds": 1518.206297563971, "env_steps_per_second": 10538.752227330835, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.68689588618219, "eta_seconds": 6072.825190255884, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0004219687363001545, "chaser_grad_norm": 0.30717000365257263, "chaser_new_fields": 76.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 5.170001029968262, "chaser_subspace_norm": 53.62324142456055, "collisions_per_episode": 1.875, "entropy": 0.14836888760328293, "explorer_grad_norm": 1.0158026963472366, "explorer_new_fields": 71.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 56.462501525878906, "explorer_subspace_norm": 14.509772300720215, "final_distance": 4.509415626525879, "max_param_abs": 0.6352270841598511, "policy_loss": -0.0007131968901994412, "value_loss": 3.9390207529067993}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:09:17.947642+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986957.947656, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.47435981461138, "elapsed_seconds": 1707.695332600968, "env_steps_per_second": 10540.521869661869, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6339060281058644, "eta_seconds": 5882.061701181112, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00013366551551963202, "chaser_grad_norm": 0.26913563534617424, "chaser_new_fields": 80.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.030000686645508, "chaser_subspace_norm": 53.20404052734375, "collisions_per_episode": 1.9250000715255737, "entropy": 0.150040864944458, "explorer_grad_norm": 0.3605656549334526, "explorer_new_fields": 72.5, "explorer_partner_vision": 0.0, "explorer_return": 58.0, "explorer_subspace_norm": 15.171175956726074, "final_distance": 5.785335063934326, "max_param_abs": 0.6562157869338989, "policy_loss": -0.001029401379868844, "value_loss": 3.9767923951148987}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:27.436674+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987147.436686, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.52638350006277, "elapsed_seconds": 1894.980600225972, "env_steps_per_second": 10554.197756755424, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.57915050018831, "eta_seconds": 5684.941800677916, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": -9.565116899454967e-05, "chaser_grad_norm": 0.08904262911528349, "chaser_new_fields": 78.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 5.560000896453857, "chaser_subspace_norm": 59.52060317993164, "collisions_per_episode": 2.950000047683716, "entropy": 0.16490384191274643, "explorer_grad_norm": 1.175610214471817, "explorer_new_fields": 69.25, "explorer_partner_vision": 0.0, "explorer_return": 52.712501525878906, "explorer_subspace_norm": 15.08913516998291, "final_distance": 5.539186000823975, "max_param_abs": 0.6929622888565063, "policy_loss": -0.0007850444589863059, "value_loss": 4.011410117149353}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:34.721943+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987334.721957, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.577963672397212, "elapsed_seconds": 2080.669220629963, "env_steps_per_second": 10573.521144960789, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5237224090471952, "eta_seconds": 5485.400672569903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0007548602515239367, "chaser_grad_norm": 0.09605849161744118, "chaser_new_fields": 79.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.750000953674316, "chaser_subspace_norm": 61.44573211669922, "collisions_per_episode": 2.4000000953674316, "entropy": 0.11969606578350067, "explorer_grad_norm": 0.5191402435302734, "explorer_new_fields": 72.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.400001525878906, "explorer_subspace_norm": 15.900866508483887, "final_distance": 4.859394550323486, "max_param_abs": 0.7189522385597229, "policy_loss": -0.0008497140859304864, "value_loss": 3.8079975247383118}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:18:40.410562+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987520.4105732, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.630036991572512, "elapsed_seconds": 2268.133169661043, "env_steps_per_second": 10581.38927688564, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4700863136691946, "eta_seconds": 5292.310729209101, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003492922228218731, "chaser_grad_norm": 0.23090405389666557, "chaser_new_fields": 79.75, "chaser_partner_vision": 0.0, "chaser_return": 5.835000514984131, "chaser_subspace_norm": 65.1114501953125, "collisions_per_episode": 1.5500000715255737, "entropy": 0.10585096292197704, "explorer_grad_norm": 1.6048751920461655, "explorer_new_fields": 72.875, "explorer_partner_vision": 0.0, "explorer_return": 58.67499923706055, "explorer_subspace_norm": 17.080568313598633, "final_distance": 4.764224052429199, "max_param_abs": 0.7304921746253967, "policy_loss": -0.0005887645710025424, "value_loss": 3.802821159362793}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:21:47.874513+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987707.8745272, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6834426114222151, "elapsed_seconds": 2460.3934011199744, "env_steps_per_second": 10567.415758863914, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4194577314153698, "eta_seconds": 5110.047833095331, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0001630648640059129, "chaser_grad_norm": 0.3192727528512478, "chaser_new_fields": 75.25, "chaser_partner_vision": 0.0, "chaser_return": 4.940001010894775, "chaser_subspace_norm": 70.8116455078125, "collisions_per_episode": 1.725000023841858, "entropy": 0.10834075324237347, "explorer_grad_norm": 0.4530724212527275, "explorer_new_fields": 69.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 53.95000076293945, "explorer_subspace_norm": 18.841650009155273, "final_distance": 5.2702436447143555, "max_param_abs": 0.756130039691925, "policy_loss": -0.0011925603797422468, "value_loss": 4.034191131591797}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:25:00.134744+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987900.1347587, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.736789553238066, "elapsed_seconds": 2652.4423916570377, "env_steps_per_second": 10556.308437864996, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3683234560135513, "eta_seconds": 4925.9644416487845, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -1.178524566994299e-05, "chaser_grad_norm": 0.08536202739924192, "chaser_new_fields": 80.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.070001125335693, "chaser_subspace_norm": 72.14111328125, "collisions_per_episode": 2.950000047683716, "entropy": 0.07558664493262768, "explorer_grad_norm": 1.3889762535691261, "explorer_new_fields": 73.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 59.57500076293945, "explorer_subspace_norm": 19.825408935546875, "final_distance": 4.625319004058838, "max_param_abs": 0.7901831269264221, "policy_loss": -0.0005326826543381458, "value_loss": 3.706700026988983}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:28:12.183735+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988092.18375, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7897381441080426, "elapsed_seconds": 2843.057318788953, "env_steps_per_second": 10552.020812854731, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3162302401800707, "eta_seconds": 4738.428864648255, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -6.961048535422676e-05, "chaser_grad_norm": 0.0827492754906416, "chaser_new_fields": 82.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.2950005531311035, "chaser_subspace_norm": 73.0795669555664, "collisions_per_episode": 2.625, "entropy": 0.07588233426213264, "explorer_grad_norm": 0.9824520200490952, "explorer_new_fields": 74.0, "explorer_partner_vision": 0.0, "explorer_return": 60.28750228881836, "explorer_subspace_norm": 21.04091453552246, "final_distance": 5.253453731536865, "max_param_abs": 0.8602654933929443, "policy_loss": -0.0011295563750906013, "value_loss": 3.637231469154358}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:31:22.798661+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988282.7986712, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8429694894869398, "elapsed_seconds": 3034.6901621529832, "env_steps_per_second": 10544.733824588327, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2644542342304097, "eta_seconds": 4552.035243229475, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00011072374704122019, "chaser_grad_norm": 0.26978185772895813, "chaser_new_fields": 81.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.12000036239624, "chaser_subspace_norm": 71.34304809570312, "collisions_per_episode": 2.174999952316284, "entropy": 0.07842518389225006, "explorer_grad_norm": 1.2386783137917519, "explorer_new_fields": 71.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 56.53750228881836, "explorer_subspace_norm": 21.640764236450195, "final_distance": 5.692413806915283, "max_param_abs": 0.8790659308433533, "policy_loss": -0.0005698082482012978, "value_loss": 3.4994654059410095}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:34:34.431505+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988474.4315186, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8960425563605451, "elapsed_seconds": 3225.753202897962, "env_steps_per_second": 10540.173987723232, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.212292870370149, "eta_seconds": 4364.254333332537, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0006385936848774948, "chaser_grad_norm": 0.06312345620244741, "chaser_new_fields": 74.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 4.77000093460083, "chaser_subspace_norm": 69.24561309814453, "collisions_per_episode": 2.174999952316284, "entropy": 0.09743076749145985, "explorer_grad_norm": 0.8409060165286064, "explorer_new_fields": 74.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 60.85000228881836, "explorer_subspace_norm": 21.00934600830078, "final_distance": 5.212314128875732, "max_param_abs": 0.9214445352554321, "policy_loss": -0.0010827581265431263, "value_loss": 3.82382869720459}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:37:45.494545+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988665.49456, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9495273919380594, "elapsed_seconds": 3418.2986109770136, "env_steps_per_second": 10531.555050338486, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.160533479035406, "eta_seconds": 4177.920524527462, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00045296810712930347, "chaser_grad_norm": 0.06943456176668406, "chaser_new_fields": 76.875, "chaser_partner_vision": 0.0, "chaser_return": 5.225000858306885, "chaser_subspace_norm": 67.38529968261719, "collisions_per_episode": 3.075000047683716, "entropy": 0.07714617811143398, "explorer_grad_norm": 0.7594472616910934, "explorer_new_fields": 72.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 57.4375, "explorer_subspace_norm": 21.665800094604492, "final_distance": 5.104313850402832, "max_param_abs": 0.9448150396347046, "policy_loss": -0.0007460070322267143, "value_loss": 3.8899757266044617}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:40:58.039955+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988858.0399704, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0030375532597342, "elapsed_seconds": 3610.935191735043, "env_steps_per_second": 10523.589591687221, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.108620453602864, "eta_seconds": 3991.033632970311, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0003543079624973317, "chaser_grad_norm": 0.3590007722377777, "chaser_new_fields": 81.5, "chaser_partner_vision": 0.0, "chaser_return": 6.210000514984131, "chaser_subspace_norm": 69.63457489013672, "collisions_per_episode": 1.8250000476837158, "entropy": 0.10340671241283417, "explorer_grad_norm": 0.45313553512096405, "explorer_new_fields": 69.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 53.76250076293945, "explorer_subspace_norm": 20.971323013305664, "final_distance": 3.7636868953704834, "max_param_abs": 0.9572228789329529, "policy_loss": -0.0009731191467725608, "value_loss": 3.7248770594596863}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:44:10.676533+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989050.6765442, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0564514804786045, "elapsed_seconds": 3803.2253297229763, "env_steps_per_second": 10517.388934963148, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0564514804786045, "eta_seconds": 3803.2253297229763, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005023974583054924, "chaser_grad_norm": 0.0713464142754674, "chaser_new_fields": 82.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.3650007247924805, "chaser_subspace_norm": 74.02313995361328, "collisions_per_episode": 2.5, "entropy": 0.09594393149018288, "explorer_grad_norm": 0.7574014440178871, "explorer_new_fields": 65.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 46.900001525878906, "explorer_subspace_norm": 21.619592666625977, "final_distance": 4.342461585998535, "max_param_abs": 1.0027084350585938, "policy_loss": -0.0005209933587471127, "value_loss": 3.910989761352539}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:47:22.966673+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989242.9666867, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1093991075366567, "elapsed_seconds": 3993.8367871319642, "env_steps_per_second": 10516.203400029486, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0037420496760228, "eta_seconds": 3613.4713788336817, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00047749529218296324, "chaser_grad_norm": 0.050129410810768604, "chaser_new_fields": 82.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.325000762939453, "chaser_subspace_norm": 84.86886596679688, "collisions_per_episode": 2.875, "entropy": 0.13744168728590012, "explorer_grad_norm": 0.6992819383740425, "explorer_new_fields": 67.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 50.57500076293945, "explorer_subspace_norm": 20.18269920349121, "final_distance": 4.5858073234558105, "max_param_abs": 1.01456880569458, "policy_loss": -0.0007293878850163082, "value_loss": 4.059099912643433}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:33.578129+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989433.5781388, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1616730993794369, "elapsed_seconds": 4182.023157765972, "env_steps_per_second": 10521.223422278872, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9504598085831756, "eta_seconds": 3421.655310899432, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00017239016186820777, "chaser_grad_norm": 0.10877509601414204, "chaser_new_fields": 82.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.385000705718994, "chaser_subspace_norm": 87.1500015258789, "collisions_per_episode": 3.0250000953674316, "entropy": 0.09259440191090107, "explorer_grad_norm": 0.4378916695713997, "explorer_new_fields": 71.0, "explorer_partner_vision": 0.0, "explorer_return": 55.525001525878906, "explorer_subspace_norm": 20.32544708251953, "final_distance": 4.560395240783691, "max_param_abs": 1.0344716310501099, "policy_loss": -0.0005659367144108529, "value_loss": 3.8039085268974304}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:53:41.764500+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989621.764515, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2148944011339011, "elapsed_seconds": 4373.619844082044, "env_steps_per_second": 10517.60364180776, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8979654269250574, "eta_seconds": 3232.675536930207, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00024058397385816332, "chaser_grad_norm": 0.08799209631979465, "chaser_new_fields": 84.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.755000591278076, "chaser_subspace_norm": 84.67320251464844, "collisions_per_episode": 2.6000001430511475, "entropy": 0.10308972001075745, "explorer_grad_norm": 0.695404440164566, "explorer_new_fields": 72.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.212501525878906, "explorer_subspace_norm": 19.738372802734375, "final_distance": 5.04566764831543, "max_param_abs": 1.058010458946228, "policy_loss": -0.000497502121493909, "value_loss": 3.8089088797569275}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:56:53.361187+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989813.3612003, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2674803307491755, "elapsed_seconds": 4562.929190697032, "env_steps_per_second": 10519.558378828915, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8449868871661169, "eta_seconds": 3041.952793798021, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 5.00601837588599e-05, "chaser_grad_norm": 0.0980833861976862, "chaser_new_fields": 84.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.650001525878906, "chaser_subspace_norm": 82.8687515258789, "collisions_per_episode": 2.674999952316284, "entropy": 0.09760069102048874, "explorer_grad_norm": 0.4400395154953003, "explorer_new_fields": 75.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 61.86249923706055, "explorer_subspace_norm": 20.37285804748535, "final_distance": 3.694981336593628, "max_param_abs": 1.099610686302185, "policy_loss": -0.0009946469450312634, "value_loss": 3.7109416127204895}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:00:02.670534+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990002.6705487, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3209165708233357, "elapsed_seconds": 4755.299654964008, "env_steps_per_second": 10514.58449054951, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7925499424940013, "eta_seconds": 2853.1797929784047, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 5.340215523474967e-05, "chaser_grad_norm": 0.0911716390401125, "chaser_new_fields": 82.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.310000896453857, "chaser_subspace_norm": 86.7227554321289, "collisions_per_episode": 2.674999952316284, "entropy": 0.10426286607980728, "explorer_grad_norm": 0.5833238512277603, "explorer_new_fields": 72.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 57.32500076293945, "explorer_subspace_norm": 19.86592674255371, "final_distance": 5.705621242523193, "max_param_abs": 1.1247121095657349, "policy_loss": -0.0009158482353655018, "value_loss": 3.847261965274811}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:03:15.040996+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990195.0410073, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3726789449461245, "elapsed_seconds": 4941.644201806048, "env_steps_per_second": 10522.813435454396, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7391348165094516, "eta_seconds": 2660.885339434026, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00010569480311406476, "chaser_grad_norm": 0.0554772624745965, "chaser_new_fields": 85.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.880001068115234, "chaser_subspace_norm": 96.93601989746094, "collisions_per_episode": 2.799999952316284, "entropy": 0.12039652466773987, "explorer_grad_norm": 0.6594202294945717, "explorer_new_fields": 69.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 53.95000076293945, "explorer_subspace_norm": 22.130189895629883, "final_distance": 5.244517803192139, "max_param_abs": 1.1261992454528809, "policy_loss": -0.0011082613698674137, "value_loss": 4.037277936935425}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:06:21.385546+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990381.3855612, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.424792619457779, "elapsed_seconds": 5129.253430048004, "env_steps_per_second": 10527.84790933885, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.686011261220412, "eta_seconds": 2469.640540393483, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -2.459619524633125e-05, "chaser_grad_norm": 0.12052098475396633, "chaser_new_fields": 84.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.67000150680542, "chaser_subspace_norm": 88.23664093017578, "collisions_per_episode": 2.5, "entropy": 0.07341985777020454, "explorer_grad_norm": 0.32117247581481934, "explorer_new_fields": 73.25, "explorer_partner_vision": 0.0, "explorer_return": 59.23749923706055, "explorer_subspace_norm": 23.01495361328125, "final_distance": 5.38625955581665, "max_param_abs": 1.1495404243469238, "policy_loss": -0.0007160339904377366, "value_loss": 3.924485921859741}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:09:28.994771+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990568.9947832, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4772423556336112, "elapsed_seconds": 5318.072480281, "env_steps_per_second": 10530.131021651858, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6331038667001191, "eta_seconds": 2279.1739201204286, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0005367802943891109, "chaser_grad_norm": 0.11330214142799377, "chaser_new_fields": 81.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.140000820159912, "chaser_subspace_norm": 84.03327941894531, "collisions_per_episode": 2.450000047683716, "entropy": 0.0882928092032671, "explorer_grad_norm": 0.7358473464846611, "explorer_new_fields": 72.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 58.375, "explorer_subspace_norm": 22.129674911499023, "final_distance": 5.254754543304443, "max_param_abs": 1.142058253288269, "policy_loss": -0.0007392778713271042, "value_loss": 4.0163044929504395}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:12:37.813823+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990757.8138375, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5299534250752833, "elapsed_seconds": 5507.83233027102, "env_steps_per_second": 10530.458540146961, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5803271612354522, "eta_seconds": 2089.1777804476283, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.001064164428248332, "chaser_grad_norm": 0.34254417940974236, "chaser_new_fields": 78.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 5.575000762939453, "chaser_subspace_norm": 80.18024444580078, "collisions_per_episode": 2.2250001430511475, "entropy": 0.11016763001680374, "explorer_grad_norm": 0.3570163697004318, "explorer_new_fields": 71.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 57.10000228881836, "explorer_subspace_norm": 22.62393569946289, "final_distance": 4.941577911376953, "max_param_abs": 1.1422009468078613, "policy_loss": -0.0016128625238849281, "value_loss": 4.240978002548218}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:15:47.573672+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990947.5736835, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.583513806088045, "elapsed_seconds": 5700.649701916962, "env_steps_per_second": 10525.116107348913, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5278379353626816, "eta_seconds": 1900.216567305654, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00037857163109160374, "chaser_grad_norm": 0.04386221989989281, "chaser_new_fields": 84.625, "chaser_partner_vision": 0.0, "chaser_return": 6.805000305175781, "chaser_subspace_norm": 83.12228393554688, "collisions_per_episode": 2.6000001430511475, "entropy": 0.08599387668073177, "explorer_grad_norm": 1.0269756838679314, "explorer_new_fields": 72.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.54999923706055, "explorer_subspace_norm": 23.459943771362305, "final_distance": 4.947392463684082, "max_param_abs": 1.1530553102493286, "policy_loss": -0.0007100305287188746, "value_loss": 3.87172269821167}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:19:00.391046+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991140.3910613, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6367809723486102, "elapsed_seconds": 5892.411500454997, "env_steps_per_second": 10522.007839271328, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4751944758431449, "eta_seconds": 1710.7001130353217, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0008950407631240775, "chaser_grad_norm": 0.06264673266559839, "chaser_new_fields": 87.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.335000514984131, "chaser_subspace_norm": 84.56585693359375, "collisions_per_episode": 2.6500000953674316, "entropy": 0.08360539376735687, "explorer_grad_norm": 1.2118786871433258, "explorer_new_fields": 73.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 59.087501525878906, "explorer_subspace_norm": 23.61764144897461, "final_distance": 5.419231414794922, "max_param_abs": 1.1369925737380981, "policy_loss": -0.001058182617882153, "value_loss": 3.515987277030945}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:22:12.152842+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991332.152854, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.690303765964733, "elapsed_seconds": 6085.093557473039, "env_steps_per_second": 10517.504685100901, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.42257594149118327, "eta_seconds": 1521.2733893682598, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005394970976517477, "chaser_grad_norm": 0.10713083669543266, "chaser_new_fields": 86.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.1750006675720215, "chaser_subspace_norm": 92.59393310546875, "collisions_per_episode": 2.2750000953674316, "entropy": 0.08037483878433704, "explorer_grad_norm": 0.7685186266899109, "explorer_new_fields": 73.0, "explorer_partner_vision": 0.0, "explorer_return": 58.525001525878906, "explorer_subspace_norm": 23.92860221862793, "final_distance": 5.838963031768799, "max_param_abs": 1.1814846992492676, "policy_loss": -0.0008086520471843883, "value_loss": 3.6930083632469177}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:25:24.834900+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991524.834915, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7419452232955437, "elapsed_seconds": 6271.0028038639575, "env_steps_per_second": 10524.632513213559, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3695035322142063, "eta_seconds": 1330.2127159711426, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00036763982743579296, "chaser_grad_norm": 0.12720348499715328, "chaser_new_fields": 85.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.895000457763672, "chaser_subspace_norm": 88.80613708496094, "collisions_per_episode": 2.4000000953674316, "entropy": 0.10433951579034328, "explorer_grad_norm": 0.6906011775135994, "explorer_new_fields": 72.0, "explorer_partner_vision": 0.0, "explorer_return": 57.13750076293945, "explorer_subspace_norm": 23.64000129699707, "final_distance": 5.227067470550537, "max_param_abs": 1.2138091325759888, "policy_loss": -0.0009032769744621305, "value_loss": 3.8834230303764343}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:28:30.744145+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991710.7441564, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7937214306763942, "elapsed_seconds": 6457.397150435019, "env_steps_per_second": 10530.558739974513, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3165390760017166, "eta_seconds": 1139.5406736061798, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 7.058613239594239e-05, "chaser_grad_norm": 0.07825290597975254, "chaser_new_fields": 84.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.840001106262207, "chaser_subspace_norm": 91.74230194091797, "collisions_per_episode": 2.0250000953674316, "entropy": 0.07394782453775406, "explorer_grad_norm": 0.4464932233095169, "explorer_new_fields": 74.5, "explorer_partner_vision": 0.0, "explorer_return": 60.88750076293945, "explorer_subspace_norm": 24.9027042388916, "final_distance": 5.423196315765381, "max_param_abs": 1.2099260091781616, "policy_loss": -0.0002897315960295188, "value_loss": 3.668298363685608}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:31:37.138494+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991897.1385074, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8463650354727805, "elapsed_seconds": 6646.91412770201, "env_steps_per_second": 10531.202698747757, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2637664336389687, "eta_seconds": 949.5591611002872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": -0.00027185974316125405, "chaser_grad_norm": 0.10700738616287708, "chaser_new_fields": 83.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.560000896453857, "chaser_subspace_norm": 91.62149810791016, "collisions_per_episode": 3.325000047683716, "entropy": 0.06692841276526451, "explorer_grad_norm": 0.49567416310310364, "explorer_new_fields": 72.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.212501525878906, "explorer_subspace_norm": 25.3548583984375, "final_distance": 5.637239933013916, "max_param_abs": 1.201588749885559, "policy_loss": -0.0006618107486282554, "value_loss": 3.7086116075515747}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:34:46.655469+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992086.6554804, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.8987362334885984, "elapsed_seconds": 6835.450440558954, "env_steps_per_second": 10533.321926054716, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21097069260984427, "eta_seconds": 759.4944933954393, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0002394027775906929, "chaser_grad_norm": 0.056074462831020355, "chaser_new_fields": 84.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.8550004959106445, "chaser_subspace_norm": 94.43164825439453, "collisions_per_episode": 2.049999952316284, "entropy": 0.03727652784436941, "explorer_grad_norm": 0.9314629286527634, "explorer_new_fields": 77.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 64.7125015258789, "explorer_subspace_norm": 25.41107177734375, "final_distance": 5.8040771484375, "max_param_abs": 1.1969293355941772, "policy_loss": -0.0005335288614745082, "value_loss": 3.7140985131263733}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:37:55.191783+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992275.1917968, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9517622308113884, "elapsed_seconds": 7026.344030920998, "env_steps_per_second": 10531.792874693647, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15825099168740983, "eta_seconds": 569.7035700746754, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": -5.4450784982947725e-05, "chaser_grad_norm": 0.15528811514377594, "chaser_new_fields": 85.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.880001068115234, "chaser_subspace_norm": 92.66083526611328, "collisions_per_episode": 2.424999952316284, "entropy": 0.09375203214585781, "explorer_grad_norm": 0.8053402975201607, "explorer_new_fields": 65.25, "explorer_partner_vision": 0.0, "explorer_return": 47.04999923706055, "explorer_subspace_norm": 24.27157211303711, "final_distance": 5.643406867980957, "max_param_abs": 1.2059156894683838, "policy_loss": -0.0010684567029812797, "value_loss": 3.949061155319214}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:41:06.085372+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992466.0853832, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0044792093539017, "elapsed_seconds": 7216.125153674046, "env_steps_per_second": 10531.968110517744, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1054989057554685, "eta_seconds": 379.7960607196866, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 8.436286475902582e-06, "chaser_grad_norm": 0.05011006910353899, "chaser_new_fields": 84.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 6.765000820159912, "chaser_subspace_norm": 93.73489379882812, "collisions_per_episode": 1.9250000715255737, "entropy": 0.07041027396917343, "explorer_grad_norm": 0.4929050877690315, "explorer_new_fields": 72.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 57.85000228881836, "explorer_subspace_norm": 25.89957618713379, "final_distance": 5.451398849487305, "max_param_abs": 1.253115177154541, "policy_loss": -0.0006270264031478234, "value_loss": 3.704754948616028}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:44:15.866496+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992655.8665109, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.056851969050541, "elapsed_seconds": 7404.667088581948, "env_steps_per_second": 10533.896941872861, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.052739794078219, "eta_seconds": 189.8632586815884, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -6.837733680590966e-05, "chaser_grad_norm": 0.0484940716996789, "chaser_new_fields": 84.25, "chaser_partner_vision": 0.0, "chaser_return": 6.725000858306885, "chaser_subspace_norm": 99.02632904052734, "collisions_per_episode": 2.049999952316284, "entropy": 0.037345570512115955, "explorer_grad_norm": 0.4485684335231781, "explorer_new_fields": 73.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 59.6875, "explorer_subspace_norm": 26.7320499420166, "final_distance": 5.3308424949646, "max_param_abs": 1.2719192504882812, "policy_loss": -0.00024499184370441185, "value_loss": 3.7726850509643555}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:47:24.408430+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992844.408441, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1029873436099864, "elapsed_seconds": 7570.7544369959505, "env_steps_per_second": 10566.978583939348, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:50:10.495778+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993010.495788, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.102990282865293, "elapsed_seconds": 7570.7650183150545, "env_steps_per_second": 10566.96381494677, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:50:10.506360+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993010.5063717, "update": 20000} diff --git a/non_social/seed_0009/attempt_01/status.json b/non_social/seed_0009/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..a18afa94c7cd00fd4fd781c814ec2f6ce2b119fe --- /dev/null +++ b/non_social/seed_0009/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.102990282865293, + "elapsed_seconds": 7570.7650183150545, + "env_steps_per_second": 10566.96381494677, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.00014854547405664803, + "chaser_grad_norm": 0.09859942272305489, + "chaser_new_fields": 82.5999984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.340001106262207, + "chaser_subspace_norm": 98.81183624267578, + "collisions_per_episode": 2.7250001430511475, + "entropy": 0.07547229155898094, + "explorer_grad_norm": 0.32836734503507614, + "explorer_new_fields": 65.67500305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 47.462501525878906, + "explorer_subspace_norm": 26.826025009155273, + "final_distance": 5.683727741241455, + "max_param_abs": 1.2476426362991333, + "policy_loss": -0.0011177757633866747, + "value_loss": 3.991724371910095 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:10.506360+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782993010.5063717, + "update": 20000 +} diff --git a/paper_random_eval.jsonl b/paper_random_eval.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..114009bf9cf9b706dc7fd41031481062b7cac1b1 --- /dev/null +++ b/paper_random_eval.jsonl @@ -0,0 +1,40 @@ +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.620987+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.200304985046387, "chaser_new_fields": 80.66999816894531, "chaser_partner_vision": 0.0, "chaser_return": 5.999996185302734, "chaser_stuck_run_steps": 0.019999999552965164, "collisions_per_episode": 2.3999998569488525, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 32.79999923706055, "explorer_partner_vision": 0.0, "explorer_return": -1.3700000047683716, "explorer_stuck_run_steps": 2.669999837875366, "final_distance": 5.008528709411621, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002434901476933704, "chaser_grad_norm": 0.07690595369786024, "chaser_new_fields": 83.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.540000915527344, "chaser_subspace_norm": 75.278076171875, "collisions_per_episode": 1.25, "entropy": 0.03620169032365084, "explorer_grad_norm": 0.2844117097556591, "explorer_new_fields": 75.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 62.76250076293945, "explorer_subspace_norm": 28.751102447509766, "final_distance": 5.306502342224121, "max_param_abs": 1.7384926080703735, "policy_loss": -0.0007488939012390006, "value_loss": 3.4187601804733276}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.718185+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.077332973480225, "chaser_new_fields": 33.65999984741211, "chaser_partner_vision": 0.0, "chaser_return": -3.331998825073242, "chaser_stuck_run_steps": 2.700000047683716, "collisions_per_episode": 2.2699999809265137, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 74.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 61.10499954223633, "explorer_stuck_run_steps": 0.07999999821186066, "final_distance": 5.209523677825928, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.818944+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.202634334564209, "chaser_new_fields": 78.47000122070312, "chaser_partner_vision": 0.0, "chaser_return": 5.575997352600098, "chaser_stuck_run_steps": 0.019999999552965164, "collisions_per_episode": 2.309999942779541, "degenerate_episodes": 83, "degenerate_fraction": 0.83, "episodes": 100, "explorer_new_fields": 32.97999954223633, "explorer_partner_vision": 0.0, "explorer_return": -1.0699999332427979, "explorer_stuck_run_steps": 2.7100000381469727, "final_distance": 5.011303424835205, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0007710583850932262, "chaser_grad_norm": 0.11076641827821732, "chaser_new_fields": 83.25, "chaser_partner_vision": 0.0, "chaser_return": 6.595000743865967, "chaser_subspace_norm": 107.82763671875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.06922571919858456, "explorer_grad_norm": 0.3988708183169365, "explorer_new_fields": 73.5, "explorer_partner_vision": 0.0, "explorer_return": 59.76250076293945, "explorer_subspace_norm": 27.074764251708984, "final_distance": 4.443612575531006, "max_param_abs": 1.3260666131973267, "policy_loss": -0.0006433645908252572, "value_loss": 3.4030256271362305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:42.915491+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.171538352966309, "chaser_new_fields": 33.95000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.2539987564086914, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 2.0899999141693115, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 71.5, "explorer_partner_vision": 0.0, "explorer_return": 56.59000015258789, "explorer_stuck_run_steps": 0.3100000023841858, "final_distance": 4.9595184326171875, "same_state_run_steps": 0.05999999865889549}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.015998+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.099634170532227, "chaser_new_fields": 81.38999938964844, "chaser_partner_vision": 0.0, "chaser_return": 6.113996505737305, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.049999952316284, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 32.849998474121094, "explorer_partner_vision": 0.0, "explorer_return": -1.3849999904632568, "explorer_stuck_run_steps": 2.700000047683716, "final_distance": 4.918816566467285, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00015983218477721994, "chaser_grad_norm": 0.06889200676232576, "chaser_new_fields": 84.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.7600016593933105, "chaser_subspace_norm": 98.9588623046875, "collisions_per_episode": 2.075000047683716, "entropy": 0.02094564540311694, "explorer_grad_norm": 1.1887003779411316, "explorer_new_fields": 82.75, "explorer_partner_vision": 0.0, "explorer_return": 73.30000305175781, "explorer_subspace_norm": 38.70703887939453, "final_distance": 4.667576789855957, "max_param_abs": 1.5199339389801025, "policy_loss": -0.0006238509232616707, "value_loss": 3.509905695915222}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.112222+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.199133396148682, "chaser_new_fields": 33.43000030517578, "chaser_partner_vision": 0.0, "chaser_return": -3.3739986419677734, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 2.049999952316284, "degenerate_episodes": 85, "degenerate_fraction": 0.85, "episodes": 100, "explorer_new_fields": 80.68999481201172, "explorer_partner_vision": 0.0, "explorer_return": 70.2699966430664, "explorer_stuck_run_steps": 0.7799999713897705, "final_distance": 4.922343730926514, "same_state_run_steps": 0.08999999612569809}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.212404+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.134275913238525, "chaser_new_fields": 86.43999481201172, "chaser_partner_vision": 0.0, "chaser_return": 7.12199592590332, "chaser_stuck_run_steps": 0.019999999552965164, "collisions_per_episode": 2.6399998664855957, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 32.790000915527344, "explorer_partner_vision": 0.0, "explorer_return": -1.2799999713897705, "explorer_stuck_run_steps": 2.6499998569488525, "final_distance": 4.301638603210449, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0002841827333717878, "chaser_grad_norm": 0.04017048981040716, "chaser_new_fields": 87.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.300001621246338, "chaser_subspace_norm": 112.96394348144531, "collisions_per_episode": 1.725000023841858, "entropy": 0.024153232108801603, "explorer_grad_norm": 0.8466866165399551, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 62.125, "explorer_subspace_norm": 36.111541748046875, "final_distance": 4.144174098968506, "max_param_abs": 1.6294050216674805, "policy_loss": -0.0005663236785924397, "value_loss": 3.4905534386634827}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.308901+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.156155109405518, "chaser_new_fields": 33.97999954223633, "chaser_partner_vision": 0.0, "chaser_return": -3.2619988918304443, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 1.8600000143051147, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 74.72000122070312, "explorer_partner_vision": 0.0, "explorer_return": 61.44999694824219, "explorer_stuck_run_steps": 0.05999999865889549, "final_distance": 5.338278293609619, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.409994+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.038296699523926, "chaser_new_fields": 78.72000122070312, "chaser_partner_vision": 0.0, "chaser_return": 5.573996543884277, "chaser_stuck_run_steps": 0.04999999701976776, "collisions_per_episode": 2.7899999618530273, "degenerate_episodes": 85, "degenerate_fraction": 0.85, "episodes": 100, "explorer_new_fields": 32.939998626708984, "explorer_partner_vision": 0.0, "explorer_return": -1.25, "explorer_stuck_run_steps": 2.700000047683716, "final_distance": 4.791200637817383, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00037036755928841064, "chaser_grad_norm": 0.08058446925133467, "chaser_new_fields": 82.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 6.330001354217529, "chaser_subspace_norm": 92.44800567626953, "collisions_per_episode": 1.8250000476837158, "entropy": 0.04575740918517113, "explorer_grad_norm": 0.3279200941324234, "explorer_new_fields": 77.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 66.0999984741211, "explorer_subspace_norm": 29.62271499633789, "final_distance": 5.308696269989014, "max_param_abs": 1.3726928234100342, "policy_loss": -0.0006299224167634065, "value_loss": 3.322889506816864}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.506454+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.115623474121094, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.0, "chaser_return": -3.289998769760132, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 2.919999837875366, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 76.08000183105469, "explorer_partner_vision": 0.0, "explorer_return": 63.32499694824219, "explorer_stuck_run_steps": 0.07000000029802322, "final_distance": 5.588272571563721, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.606931+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.116372585296631, "chaser_new_fields": 76.8699951171875, "chaser_partner_vision": 0.0, "chaser_return": 5.221996784210205, "chaser_stuck_run_steps": 0.9899999499320984, "collisions_per_episode": 2.9099998474121094, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 32.80999755859375, "explorer_partner_vision": 0.0, "explorer_return": -1.5199999809265137, "explorer_stuck_run_steps": 2.7300000190734863, "final_distance": 4.800493240356445, "same_state_run_steps": 0.09999999403953552}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00022429173202174373, "chaser_grad_norm": 0.04249562695622444, "chaser_new_fields": 80.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.015000820159912, "chaser_subspace_norm": 93.73216247558594, "collisions_per_episode": 1.024999976158142, "entropy": 0.05529557354748249, "explorer_grad_norm": 0.39152009785175323, "explorer_new_fields": 73.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 59.42499923706055, "explorer_subspace_norm": 117.07512664794922, "final_distance": 4.730400562286377, "max_param_abs": 1.606895923614502, "policy_loss": -0.0002411644566997495, "value_loss": 3.523293375968933}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.703008+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.121584415435791, "chaser_new_fields": 33.97999954223633, "chaser_partner_vision": 0.0, "chaser_return": -3.2599987983703613, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 2.3499999046325684, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 71.93999481201172, "explorer_partner_vision": 0.0, "explorer_return": 57.11499786376953, "explorer_stuck_run_steps": 1.559999942779541, "final_distance": 4.933297634124756, "same_state_run_steps": 0.11999999731779099}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.802941+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.091162204742432, "chaser_new_fields": 77.73999786376953, "chaser_partner_vision": 0.0, "chaser_return": 5.407996654510498, "chaser_stuck_run_steps": 0.09999999403953552, "collisions_per_episode": 2.509999990463257, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 32.80999755859375, "explorer_partner_vision": 0.0, "explorer_return": -1.3399999141693115, "explorer_stuck_run_steps": 2.700000047683716, "final_distance": 5.153555870056152, "same_state_run_steps": 0.019999999552965164}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 8.635803430578903e-06, "chaser_grad_norm": 0.10996681079268456, "chaser_new_fields": 82.375, "chaser_partner_vision": 0.0, "chaser_return": 6.390000820159912, "chaser_subspace_norm": 73.99776458740234, "collisions_per_episode": 1.149999976158142, "entropy": 0.05159799661487341, "explorer_grad_norm": 0.7565127648413181, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.0, "explorer_return": 41.23749923706055, "explorer_subspace_norm": 46.546077728271484, "final_distance": 4.668344020843506, "max_param_abs": 1.2401388883590698, "policy_loss": -0.0002673804239634592, "value_loss": 3.8410890698432922}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:43.900898+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.003658771514893, "chaser_new_fields": 33.75, "chaser_partner_vision": 0.0, "chaser_return": -3.3079988956451416, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 2.4800000190734863, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 60.68000030517578, "explorer_partner_vision": 0.0, "explorer_return": 40.404998779296875, "explorer_stuck_run_steps": 0.3199999928474426, "final_distance": 4.824435710906982, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.001633+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.128373146057129, "chaser_new_fields": 82.11000061035156, "chaser_partner_vision": 0.0, "chaser_return": 6.269996166229248, "chaser_stuck_run_steps": 0.08999999612569809, "collisions_per_episode": 2.93999981880188, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.040000915527344, "explorer_partner_vision": 0.0, "explorer_return": -0.9799999594688416, "explorer_stuck_run_steps": 2.7200000286102295, "final_distance": 5.028230667114258, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005693047922270544, "chaser_grad_norm": 0.08362068422138691, "chaser_new_fields": 82.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 6.3000006675720215, "chaser_subspace_norm": 109.71160888671875, "collisions_per_episode": 2.174999952316284, "entropy": 0.05865261983126402, "explorer_grad_norm": 0.6362885981798172, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.0, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 35.61513900756836, "final_distance": 4.870059967041016, "max_param_abs": 1.4202895164489746, "policy_loss": -0.0008471746835780891, "value_loss": 4.014994144439697}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.097744+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.059652328491211, "chaser_new_fields": 33.790000915527344, "chaser_partner_vision": 0.0, "chaser_return": -3.2999987602233887, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 2.2899999618530273, "degenerate_episodes": 90, "degenerate_fraction": 0.9, "episodes": 100, "explorer_new_fields": 61.14999771118164, "explorer_partner_vision": 0.0, "explorer_return": 41.125, "explorer_stuck_run_steps": 0.19999998807907104, "final_distance": 4.463276386260986, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.197891+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.262848854064941, "chaser_new_fields": 78.5, "chaser_partner_vision": 0.0, "chaser_return": 5.547996520996094, "chaser_stuck_run_steps": 0.03999999910593033, "collisions_per_episode": 2.440000057220459, "degenerate_episodes": 83, "degenerate_fraction": 0.83, "episodes": 100, "explorer_new_fields": 32.88999938964844, "explorer_partner_vision": 0.0, "explorer_return": -1.3700000047683716, "explorer_stuck_run_steps": 2.740000009536743, "final_distance": 5.147022724151611, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 3.927184380664306e-06, "chaser_grad_norm": 0.07202051393687725, "chaser_new_fields": 81.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.145000457763672, "chaser_subspace_norm": 132.7053680419922, "collisions_per_episode": 1.975000023841858, "entropy": 0.13418130949139595, "explorer_grad_norm": 0.6797178164124489, "explorer_new_fields": 68.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 52.48749923706055, "explorer_subspace_norm": 61.86991500854492, "final_distance": 4.521096229553223, "max_param_abs": 1.4186153411865234, "policy_loss": -0.0009259324722812323, "value_loss": 4.002953112125397}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.294102+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.165109157562256, "chaser_new_fields": 33.48999786376953, "chaser_partner_vision": 0.0, "chaser_return": -3.343998908996582, "chaser_stuck_run_steps": 2.5799999237060547, "collisions_per_episode": 2.2699999809265137, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 68.97000122070312, "explorer_partner_vision": 0.0, "explorer_return": 52.70499801635742, "explorer_stuck_run_steps": 0.8499999642372131, "final_distance": 5.33642578125, "same_state_run_steps": 0.12999999523162842}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.394659+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.134922504425049, "chaser_new_fields": 82.80999755859375, "chaser_partner_vision": 0.0, "chaser_return": 6.391995906829834, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 2.819999933242798, "degenerate_episodes": 83, "degenerate_fraction": 0.83, "episodes": 100, "explorer_new_fields": 32.959999084472656, "explorer_partner_vision": 0.0, "explorer_return": -1.2649999856948853, "explorer_stuck_run_steps": 2.679999828338623, "final_distance": 5.392402648925781, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "none", "task": "non_social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00014854547405664803, "chaser_grad_norm": 0.09859942272305489, "chaser_new_fields": 82.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 6.340001106262207, "chaser_subspace_norm": 98.81183624267578, "collisions_per_episode": 2.7250001430511475, "entropy": 0.07547229155898094, "explorer_grad_norm": 0.32836734503507614, "explorer_new_fields": 65.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 26.826025009155273, "final_distance": 5.683727741241455, "max_param_abs": 1.2476426362991333, "policy_loss": -0.0011177757633866747, "value_loss": 3.991724371910095}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.491290+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 4.98706579208374, "chaser_new_fields": 33.81999969482422, "chaser_partner_vision": 0.0, "chaser_return": -3.3119988441467285, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 2.490000009536743, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 64.93999481201172, "explorer_partner_vision": 0.0, "explorer_return": 46.67499923706055, "explorer_stuck_run_steps": 0.029999999329447746, "final_distance": 4.802732467651367, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.600352+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.9226698875427246, "chaser_new_fields": 31.739999771118164, "chaser_partner_vision": 0.9387999773025513, "chaser_return": 22.954994201660156, "chaser_stuck_run_steps": 0.6100000143051147, "collisions_per_episode": 24.729999542236328, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 27.779998779296875, "explorer_partner_vision": 0.9387999773025513, "explorer_return": -26.604999542236328, "explorer_stuck_run_steps": 3.259999990463257, "final_distance": 1.7273223400115967, "same_state_run_steps": 0.26999998092651367}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.704418+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.418818473815918, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.2913999855518341, "chaser_return": -1.88499915599823, "chaser_stuck_run_steps": 2.7100000381469727, "collisions_per_episode": 1.25, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 55.38999938964844, "explorer_partner_vision": 0.2913999855518341, "explorer_return": 32.06999969482422, "explorer_stuck_run_steps": 0.12999999523162842, "final_distance": 5.25391149520874, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.812676+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 2.0339229106903076, "chaser_new_fields": 33.209999084472656, "chaser_partner_vision": 0.9183999300003052, "chaser_return": 21.927995681762695, "chaser_stuck_run_steps": 0.6699999570846558, "collisions_per_episode": 23.51999855041504, "degenerate_episodes": 92, "degenerate_fraction": 0.92, "episodes": 100, "explorer_new_fields": 26.84000015258789, "explorer_partner_vision": 0.9183999300003052, "explorer_return": -26.510000228881836, "explorer_stuck_run_steps": 3.169999837875366, "final_distance": 1.7653462886810303, "same_state_run_steps": 0.25}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:44.916857+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.981630325317383, "chaser_new_fields": 33.93000030517578, "chaser_partner_vision": 0.2247999906539917, "chaser_return": -2.4529991149902344, "chaser_stuck_run_steps": 2.690000057220459, "collisions_per_episode": 0.7099999785423279, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 53.73999786376953, "explorer_partner_vision": 0.2247999906539917, "explorer_return": 29.98499870300293, "explorer_stuck_run_steps": 0.28999999165534973, "final_distance": 5.55134916305542, "same_state_run_steps": 0.03999999910593033}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.025324+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.846373200416565, "chaser_new_fields": 30.559999465942383, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 25.47699737548828, "chaser_stuck_run_steps": 2.4099998474121094, "collisions_per_episode": 27.189998626708984, "degenerate_episodes": 95, "degenerate_fraction": 0.95, "episodes": 100, "explorer_new_fields": 25.549999237060547, "explorer_partner_vision": 0.9455000162124634, "explorer_return": -29.889999389648438, "explorer_stuck_run_steps": 3.359999895095825, "final_distance": 1.641021728515625, "same_state_run_steps": 0.5399999618530273}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.131142+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.628232479095459, "chaser_new_fields": 33.95000076293945, "chaser_partner_vision": 0.26339998841285706, "chaser_return": -2.1799991130828857, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.9599999785423279, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 54.0, "explorer_partner_vision": 0.26339998841285706, "explorer_return": 30.189998626708984, "explorer_stuck_run_steps": 0.019999999552965164, "final_distance": 5.8266167640686035, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.255109+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.8304731845855713, "chaser_new_fields": 30.989999771118164, "chaser_partner_vision": 0.9462000131607056, "chaser_return": 27.23999786376953, "chaser_stuck_run_steps": 0.85999995470047, "collisions_per_episode": 28.760000228881836, "degenerate_episodes": 98, "degenerate_fraction": 0.98, "episodes": 100, "explorer_new_fields": 25.8799991607666, "explorer_partner_vision": 0.9462000131607056, "explorer_return": -31.03499984741211, "explorer_stuck_run_steps": 3.3499999046325684, "final_distance": 1.5262326002120972, "same_state_run_steps": 0.3799999952316284}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.376361+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.968653202056885, "chaser_new_fields": 33.86000061035156, "chaser_partner_vision": 0.22360000014305115, "chaser_return": -2.4379990100860596, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 0.7400000095367432, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 55.52000045776367, "explorer_partner_vision": 0.22360000014305115, "explorer_return": 32.59499740600586, "explorer_stuck_run_steps": 0.17999999225139618, "final_distance": 5.932736873626709, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.500759+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.7290464639663696, "chaser_new_fields": 28.920000076293945, "chaser_partner_vision": 0.9477999806404114, "chaser_return": 36.003997802734375, "chaser_stuck_run_steps": 1.3799999952316284, "collisions_per_episode": 37.13999938964844, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 23.78999900817871, "explorer_partner_vision": 0.9477999806404114, "explorer_return": -39.125, "explorer_stuck_run_steps": 3.5999999046325684, "final_distance": 1.566463589668274, "same_state_run_steps": 0.5299999713897705}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.610487+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.055323123931885, "chaser_new_fields": 33.90999984741211, "chaser_partner_vision": 0.204599991440773, "chaser_return": -2.4979991912841797, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.6800000071525574, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 46.849998474121094, "explorer_partner_vision": 0.204599991440773, "explorer_return": 19.69499969482422, "explorer_stuck_run_steps": 0.07000000029802322, "final_distance": 6.047131061553955, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.725728+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 2.022212505340576, "chaser_new_fields": 31.869998931884766, "chaser_partner_vision": 0.9148999452590942, "chaser_return": 23.19499397277832, "chaser_stuck_run_steps": 0.85999995470047, "collisions_per_episode": 24.889999389648438, "degenerate_episodes": 92, "degenerate_fraction": 0.92, "episodes": 100, "explorer_new_fields": 27.619998931884766, "explorer_partner_vision": 0.9148999452590942, "explorer_return": -25.889999389648438, "explorer_stuck_run_steps": 3.240000009536743, "final_distance": 1.8503373861312866, "same_state_run_steps": 0.25999999046325684}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.829596+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.011486530303955, "chaser_new_fields": 33.73999786376953, "chaser_partner_vision": 0.2134999781847, "chaser_return": -2.4539990425109863, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 0.7400000095367432, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 43.37999725341797, "explorer_partner_vision": 0.2134999781847, "explorer_return": 14.429999351501465, "explorer_stuck_run_steps": 0.17000000178813934, "final_distance": 5.857645034790039, "same_state_run_steps": 0.019999999552965164}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:45.937143+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.8730149269104004, "chaser_new_fields": 31.01999855041504, "chaser_partner_vision": 0.939300000667572, "chaser_return": 26.81099510192871, "chaser_stuck_run_steps": 0.7799999713897705, "collisions_per_episode": 28.369998931884766, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 25.760000228881836, "explorer_partner_vision": 0.939300000667572, "explorer_return": -30.899999618530273, "explorer_stuck_run_steps": 3.3299999237060547, "final_distance": 1.75227952003479, "same_state_run_steps": 0.2800000011920929}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.040804+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.9493937492370605, "chaser_new_fields": 33.93000030517578, "chaser_partner_vision": 0.21109998226165771, "chaser_return": -2.5089991092681885, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 0.6699999570846558, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 53.27000045776367, "explorer_partner_vision": 0.21109998226165771, "explorer_return": 29.344999313354492, "explorer_stuck_run_steps": 0.07000000029802322, "final_distance": 5.721303462982178, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.157423+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.8591976165771484, "chaser_new_fields": 30.69999885559082, "chaser_partner_vision": 0.9456999897956848, "chaser_return": 26.13799285888672, "chaser_stuck_run_steps": 1.0299999713897705, "collisions_per_episode": 27.799999237060547, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 26.239999771118164, "explorer_partner_vision": 0.9456999897956848, "explorer_return": -29.90999984741211, "explorer_stuck_run_steps": 3.369999885559082, "final_distance": 1.777151107788086, "same_state_run_steps": 0.4899999797344208}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.263149+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.806032657623291, "chaser_new_fields": 33.84000015258789, "chaser_partner_vision": 0.23879998922348022, "chaser_return": -2.4399991035461426, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 0.7400000095367432, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 57.55999755859375, "explorer_partner_vision": 0.23879998922348022, "explorer_return": 35.744998931884766, "explorer_stuck_run_steps": 0.08999999612569809, "final_distance": 6.035492897033691, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.371355+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.9622390270233154, "chaser_new_fields": 31.35999870300293, "chaser_partner_vision": 0.9269000291824341, "chaser_return": 23.311994552612305, "chaser_stuck_run_steps": 0.6299999952316284, "collisions_per_episode": 25.100000381469727, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 27.68000030517578, "explorer_partner_vision": 0.9269000291824341, "explorer_return": -26.309999465942383, "explorer_stuck_run_steps": 3.1399998664855957, "final_distance": 1.7391915321350098, "same_state_run_steps": 0.20999999344348907}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.475135+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.686277866363525, "chaser_new_fields": 33.880001068115234, "chaser_partner_vision": 0.2646999955177307, "chaser_return": -2.040999412536621, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 1.090000033378601, "degenerate_episodes": 88, "degenerate_fraction": 0.88, "episodes": 100, "explorer_new_fields": 55.25, "explorer_partner_vision": 0.2646999955177307, "explorer_return": 31.864999771118164, "explorer_stuck_run_steps": 0.26999998092651367, "final_distance": 5.499191761016846, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.583483+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 2.13496994972229, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.8927000164985657, "chaser_return": 20.284996032714844, "chaser_stuck_run_steps": 0.550000011920929, "collisions_per_episode": 21.850000381469727, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 28.149999618530273, "explorer_partner_vision": 0.8927000164985657, "explorer_return": -23.260000228881836, "explorer_stuck_run_steps": 3.2200000286102295, "final_distance": 1.8080952167510986, "same_state_run_steps": 0.25}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "mlp", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "cost_per_hour": null, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "env": {"grid_size": 10, "max_steps": 100, "partner_visibility": "partial", "task": "social", "vision_radius": 3}, "experiment_id": "mouse-run-run-2-mlp", "finite_guard": true, "fused_agent_rollout": false, "gae_lambda": 0.95, "gamma": 0.99, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "log_every": 500, "metrics_path": null, "ppo_epochs": 4, "recurrent_l2_coef": 0.3, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "status_every_seconds": 1800.0, "status_path": null, "subspace_metric_period": 10, "tensorboard_dir": null, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "checkpoint_metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-mlp --output runs/mouse-run-run-2-mlp/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-02T11:50:46.687176+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 100, "error": null, "evaluation_protocol": "paper_random_opponent_v1", "experiment_id": "mouse-run-run-2-mlp", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.822822093963623, "chaser_new_fields": 33.88999938964844, "chaser_partner_vision": 0.2704000174999237, "chaser_return": -1.4439992904663086, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 1.659999966621399, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 49.029998779296875, "explorer_partner_vision": 0.2704000174999237, "explorer_return": 22.23499870300293, "explorer_stuck_run_steps": 0.14999999105930328, "final_distance": 6.008201599121094, "same_state_run_steps": 0.029999999329447746}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0009"} diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fedac4838c0a4b4b43e5192d523daee667dd435e --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:47235ca3adb8a5800cf9afc150039b2a465987c50ff85daa8b9f7e2eb91b4700 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4f0b5ad739df493227634115918497ba76bd22dc --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e6dcd1a1b640313bf009d7c3af192d2acfd254fa7e355e524c6f8ba35c4eae0a +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..6647de0c1db342b9fb773e735e15f82ce214b7d7 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6004b9a96caddd23e6c5370cb3a7dcbc79e7ccf86415bfc293f4f2859c725039 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..ccc8a4fa3e2cef111206ffb436dfd432b1959246 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a26129d1634332982cfc7436c50c3b54df91edd77569ff235c1a0d8f18507383 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..307f2ddb08849ed394299b1c9a155e952a3cabd1 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8001169fe126091e26b133edb59a13bb3bff90ab5826b6e32dd7745102385788 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a41a856aebab4e7ae4279b1e2d830d07c295e61f --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:620dbd11a3b2f5485389a71091b2f45458b317db672910f107c567a98f03c28f +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..f9ccba689a2e6016a54d3a0ab0d68fd67b761a48 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb1d4e549c50e6326007a139e6491233b34389321ae2fbab23d58825b7dbd8d4 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c029290e98019edff1d0217c2b34635747a7336f --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a941f04ea8dcce75abf0a42ed2329f1c987af0a7158fd5f92592eb67d418386e +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..670506d72dcbcee7c540a6b0b4c415ce604b97fb --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3c249dac8de670dffeaf04b791c3b76f1f7015fad519831221eaf905d9e1b668 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8af80f691a3b2dddaf5b894a1f405c9b09db97ae --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ecd6db0f030df8ca8995578bb1416a8d00c4293749b545e0dbf7e583909903d4 +size 46506651 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..1afcb3824de001dec78ce3125059060fb0c05de9 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0bad577c731d05df8b417b04bccfce215674bf0397bf0c87f6ea3f629e373f67 +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..085c503d20f82b12a97411571de3015346504d95 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:0e681e5199bb74bbb486c17420e6531bde8ca7f4873ba857ffd839b9e2e0bd1f +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c5c10f25b09eaeae7a6628802ed1c7a6412b6d87 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2c1daa497f3a0b5503f0e5f7a87fea538c6dc3407fc2adf0dfa6d26bc726d5b4 +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..636ebf54e0e9c3ecb8762abbba5ac07e230e9b9d --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ceb7f1689016b609d5e2713a0a2661ca9f83078a5aa89dd66851f17525c1f890 +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8d8b8c005e2cdd07aec7514274ca29af414716d7 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c12193945e93692204781680d7af3c920b1a0307a09e31072266eb2a5e4610db +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aa1d6fc00d0ac7ca8f09f751e21a3e77be734836 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:01465d07d01ed57f5c00a757c8a01ed8d23ccab455e35db5ae547eff48410011 +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..07460da0403efd35a6539e65d9d123bf4f56088e --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6433519f01ecece7e0023cf22dd156df34773a4064fd7e051275a0b34b41510f +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..93280def3b538bd4fd1aa331654d71a7cdf013e0 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:86ef95b85618bcb31066bdd41f692e9b045da238f8e252197743daff8da6ac17 +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..43fd2e727d855a583c642fbf689adb2933a19c3d --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:76c5e4773e0a78d321490080d05ddc09f3af942e00d6c8958cdb4e3d7cefcea4 +size 46506635 diff --git a/paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..aaa9012feec8d39721c3a0604aafa04021543438 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:e98fced65d944fcdfc904a1b8344be4caaa933f30cb055861f623183b3df5dc9 +size 46506635 diff --git a/paper_rollouts/paper_rollout_records.jsonl b/paper_rollouts/paper_rollout_records.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..2b15bf827fd1c5dc864eb0bb1b160b67c487b140 --- /dev/null +++ b/paper_rollouts/paper_rollout_records.jsonl @@ -0,0 +1,20 @@ +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:48.710464+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:49.261015+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:49.810154+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:50.371542+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:50.918340+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:51.471824+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:52.025242+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:52.569627+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:53.116716+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:53.672368+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "non_social", "unit_id": "non_social/seed_0009"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:54.273478+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0000"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:54.873436+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0001"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:55.461961+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0002"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:56.048411+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0003"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:56.645581+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0004"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:57.229171+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0005"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:57.814519+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0006"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:58.414469+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0007"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:58.996792+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0008"} +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "tensor_count": 66}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-mlp --output-root runs/mouse-run-run-2-mlp/paper_rollouts --device cuda", "created_at": "2026-07-02T11:50:59.591392+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-mlp", "max_steps": 500, "output": "runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "status": "ok", "task": "social", "unit_id": "social/seed_0009"} diff --git a/raw_records.jsonl b/raw_records.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..7c17e72b37c61d8e859423d2cb73380bc7ba2fe7 --- /dev/null +++ b/raw_records.jsonl @@ -0,0 +1,41 @@ +{"event": "runner_invocation", "experiment": "mouse-run-run-2-mlp", "payload": {"config": {"architecture": "mlp", "batch_size": 40, "checkpoint_every": 1000, "checkpoint_every_seconds": 1800.0, "clip_epsilon": 0.2, "concurrency": 10, "cuda_tf32": true, "device": "cuda", "entropy_coef": 0.01, "finite_guard": true, "fused_agent_rollout": false, "grad_clip": 1.0, "hidden_size": 256, "learning_rate": 0.0003, "max_steps": 100, "ppo_epochs": 4, "preset": "paper_text", "recurrent_l2_coef": 0.3, "status_every_seconds": 1800.0, "subspace_metric_period": 10, "triton_env_step": true, "updates": 20000, "value_clip": 10.0, "value_coef": 0.5}, "config_sha256": "1e164a3dab3a7ecde08c0595611a60ed85489e06861f477e46efeb1be25d94c3", "created_at": "2026-07-02T07:36:16.447422+00:00", "experiment": "mouse-run-run-2-mlp", "experiment_spec": "experiments/paper_marl_2026/SPEC.md", "experiment_spec_sha256": "e878bd03a06735c93c8caab4fc769ce63f026a538dfccbc87b588d432e48761e", "initial_attempts": [{"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0000_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "task": "social", "unit_id": "social/seed_0000"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0000_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "task": "non_social", "unit_id": "non_social/seed_0000"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0001_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "task": "social", "unit_id": "social/seed_0001"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0001_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "task": "non_social", "unit_id": "non_social/seed_0001"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0002_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "task": "social", "unit_id": "social/seed_0002"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0002_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "task": "non_social", "unit_id": "non_social/seed_0002"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0003_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "task": "social", "unit_id": "social/seed_0003"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0003_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "task": "non_social", "unit_id": "non_social/seed_0003"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0004_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "task": "social", "unit_id": "social/seed_0004"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0004_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "task": "non_social", "unit_id": "non_social/seed_0004"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0005_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "task": "social", "unit_id": "social/seed_0005"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0005_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "task": "non_social", "unit_id": "non_social/seed_0005"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0006_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "task": "social", "unit_id": "social/seed_0006"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0006_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "task": "non_social", "unit_id": "non_social/seed_0006"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0007_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "task": "social", "unit_id": "social/seed_0007"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0007_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "task": "non_social", "unit_id": "non_social/seed_0007"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0008_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "task": "social", "unit_id": "social/seed_0008"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0008_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "task": "non_social", "unit_id": "non_social/seed_0008"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0009_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "task": "social", "unit_id": "social/seed_0009"}, {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0009_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "task": "non_social", "unit_id": "non_social/seed_0009"}], "max_attempts": 2, "provenance": {"command": "scripts/run_local_experiment.py --config experiments/paper_marl_2026/configs/arch_mlp.json", "created_at": "2026-07-02T07:36:16.447522+00:00", "dependencies": {"huggingface_hub": "1.21.0", "numpy": "2.5.0", "safetensors": "0.8.0", "tensorboard": "2.21.0", "torch": "2.12.1", "triton": "3.7.1"}, "executable": "/root/mouse-run-run-perf/.venv/bin/python3", "git": {"branch": "main", "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", "dirty": false, "sha": "6282c650d281b31d646d537dd3f57c968806a423", "short_sha": "6282c65", "status_short": ""}, "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", "schema_version": 1, "torch": {"cuda_available": true, "cuda_device_count": 1, "cuda_devices": [{"index": 0, "major": 8, "minor": 9, "multi_processor_count": 66, "name": "NVIDIA GeForce RTX 4070 Ti SUPER", "total_memory_bytes": 16717578240}], "cuda_version": "13.0", "cudnn_version": 92000, "mps_available": false, "version": "2.12.1+cu130"}}, "schema_version": 2, "total_environment_steps": 1600000000, "total_units": 20, "units": [{"seed": 0, "task": "social", "unit_id": "social/seed_0000"}, {"seed": 0, "task": "non_social", "unit_id": "non_social/seed_0000"}, {"seed": 1, "task": "social", "unit_id": "social/seed_0001"}, {"seed": 1, "task": "non_social", "unit_id": "non_social/seed_0001"}, {"seed": 2, "task": "social", "unit_id": "social/seed_0002"}, {"seed": 2, "task": "non_social", "unit_id": "non_social/seed_0002"}, {"seed": 3, "task": "social", "unit_id": "social/seed_0003"}, {"seed": 3, "task": "non_social", "unit_id": "non_social/seed_0003"}, {"seed": 4, "task": "social", "unit_id": "social/seed_0004"}, {"seed": 4, "task": "non_social", "unit_id": "non_social/seed_0004"}, {"seed": 5, "task": "social", "unit_id": "social/seed_0005"}, {"seed": 5, "task": "non_social", "unit_id": "non_social/seed_0005"}, {"seed": 6, "task": "social", "unit_id": "social/seed_0006"}, {"seed": 6, "task": "non_social", "unit_id": "non_social/seed_0006"}, {"seed": 7, "task": "social", "unit_id": "social/seed_0007"}, {"seed": 7, "task": "non_social", "unit_id": "non_social/seed_0007"}, {"seed": 8, "task": "social", "unit_id": "social/seed_0008"}, {"seed": 8, "task": "non_social", "unit_id": "non_social/seed_0008"}, {"seed": 9, "task": "social", "unit_id": "social/seed_0009"}, {"seed": 9, "task": "non_social", "unit_id": "non_social/seed_0009"}]}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.520057+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0000_attempt_01.log", "pid": 52153, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T07:36:16.521612+00:00", "started_monotonic": 627610.995599802, "task": "social", "unit_id": "social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.521624+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0000_attempt_01.log", "pid": 52154, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T07:36:16.522457+00:00", "started_monotonic": 627610.996443721, "task": "non_social", "unit_id": "non_social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.522467+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0001_attempt_01.log", "pid": 52155, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T07:36:16.523003+00:00", "started_monotonic": 627610.996986632, "task": "social", "unit_id": "social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.523008+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0001_attempt_01.log", "pid": 52156, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T07:36:16.523439+00:00", "started_monotonic": 627610.997422065, "task": "non_social", "unit_id": "non_social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.523444+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0002_attempt_01.log", "pid": 52157, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T07:36:16.523887+00:00", "started_monotonic": 627610.997869881, "task": "social", "unit_id": "social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.523891+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0002_attempt_01.log", "pid": 52158, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T07:36:16.524309+00:00", "started_monotonic": 627610.998292259, "task": "non_social", "unit_id": "non_social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.524314+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0003_attempt_01.log", "pid": 52159, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T07:36:16.524739+00:00", "started_monotonic": 627610.998721919, "task": "social", "unit_id": "social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.524743+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0003_attempt_01.log", "pid": 52160, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T07:36:16.525156+00:00", "started_monotonic": 627610.999139673, "task": "non_social", "unit_id": "non_social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.525161+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0004_attempt_01.log", "pid": 52161, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T07:36:16.525631+00:00", "started_monotonic": 627610.999614048, "task": "social", "unit_id": "social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.525635+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0004_attempt_01.log", "pid": 52162, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T07:36:16.526501+00:00", "started_monotonic": 627611.000488049, "task": "non_social", "unit_id": "non_social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T07:36:16.526512+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0000_attempt_01.log", "pid": 52154, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T07:36:16.522457+00:00", "started_monotonic": 627610.996443721, "task": "non_social", "unit_id": "non_social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:08.363674+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0005_attempt_01.log", "pid": 52725, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T09:43:18.366889+00:00", "started_monotonic": 635232.840878297, "task": "social", "unit_id": "social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:18.366904+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0001_attempt_01.log", "pid": 52156, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T07:36:16.523439+00:00", "started_monotonic": 627610.997422065, "task": "non_social", "unit_id": "non_social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:18.525107+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0005_attempt_01.log", "pid": 52759, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T09:43:28.528215+00:00", "started_monotonic": 635243.002205786, "task": "non_social", "unit_id": "non_social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:28.528236+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0002_attempt_01.log", "pid": 52158, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T07:36:16.524309+00:00", "started_monotonic": 627610.998292259, "task": "non_social", "unit_id": "non_social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:28.587719+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0006_attempt_01.log", "pid": 52793, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T09:43:38.591042+00:00", "started_monotonic": 635253.06503163, "task": "social", "unit_id": "social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:38.591058+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0000_attempt_01.log", "pid": 52153, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T07:36:16.521612+00:00", "started_monotonic": 627610.995599802, "task": "social", "unit_id": "social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:38.602232+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0003_attempt_01.log", "pid": 52160, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T07:36:16.525156+00:00", "started_monotonic": 627610.999139673, "task": "non_social", "unit_id": "non_social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:38.614279+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0006_attempt_01.log", "pid": 52827, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T09:43:48.617806+00:00", "started_monotonic": 635263.091794424, "task": "non_social", "unit_id": "non_social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.617820+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0007_attempt_01.log", "pid": 52828, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T09:43:48.618546+00:00", "started_monotonic": 635263.092532864, "task": "social", "unit_id": "social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.618557+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0001_attempt_01.log", "pid": 52155, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T07:36:16.523003+00:00", "started_monotonic": 627610.996986632, "task": "social", "unit_id": "social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.626631+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0002_attempt_01.log", "pid": 52157, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T07:36:16.523887+00:00", "started_monotonic": 627610.997869881, "task": "social", "unit_id": "social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.634605+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0003_attempt_01.log", "pid": 52159, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T07:36:16.524739+00:00", "started_monotonic": 627610.998721919, "task": "social", "unit_id": "social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.659545+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0004_attempt_01.log", "pid": 52161, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T07:36:16.525631+00:00", "started_monotonic": 627610.999614048, "task": "social", "unit_id": "social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.667523+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0004_attempt_01.log", "pid": 52162, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T07:36:16.526501+00:00", "started_monotonic": 627611.000488049, "task": "non_social", "unit_id": "non_social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:48.678869+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0007_attempt_01.log", "pid": 52895, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T09:43:58.682512+00:00", "started_monotonic": 635273.156502974, "task": "non_social", "unit_id": "non_social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:58.682532+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0008_attempt_01.log", "pid": 52896, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T09:43:58.683379+00:00", "started_monotonic": 635273.157367747, "task": "social", "unit_id": "social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:58.683393+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0008_attempt_01.log", "pid": 52897, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T09:43:58.684210+00:00", "started_monotonic": 635273.158198384, "task": "non_social", "unit_id": "non_social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:58.684224+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0009_attempt_01.log", "pid": 52898, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T09:43:58.684978+00:00", "started_monotonic": 635273.158965726, "task": "social", "unit_id": "social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:58.684990+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0009_attempt_01.log", "pid": 52899, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T09:43:58.685755+00:00", "started_monotonic": 635273.159742929, "task": "non_social", "unit_id": "non_social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-02T09:43:58.685768+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0005_attempt_01.log", "pid": 52725, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T09:43:18.366889+00:00", "started_monotonic": 635232.840878297, "task": "social", "unit_id": "social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-02T11:49:20.888272+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0005_attempt_01.log", "pid": 52759, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T09:43:28.528215+00:00", "started_monotonic": 635243.002205786, "task": "non_social", "unit_id": "non_social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-02T11:49:40.902078+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0006_attempt_01.log", "pid": 52793, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T09:43:38.591042+00:00", "started_monotonic": 635253.06503163, "task": "social", "unit_id": "social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-02T11:49:40.909677+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0006_attempt_01.log", "pid": 52827, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T09:43:48.617806+00:00", "started_monotonic": 635263.091794424, "task": "non_social", "unit_id": "non_social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:00.923016+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0007_attempt_01.log", "pid": 52828, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T09:43:48.618546+00:00", "started_monotonic": 635263.092532864, "task": "social", "unit_id": "social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:20.933526+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0007_attempt_01.log", "pid": 52895, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T09:43:58.682512+00:00", "started_monotonic": 635273.156502974, "task": "non_social", "unit_id": "non_social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:20.938236+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0008_attempt_01.log", "pid": 52897, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T09:43:58.684210+00:00", "started_monotonic": 635273.158198384, "task": "non_social", "unit_id": "non_social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:20.942696+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0009_attempt_01.log", "pid": 52899, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T09:43:58.685755+00:00", "started_monotonic": 635273.159742929, "task": "non_social", "unit_id": "non_social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:20.949089+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0008_attempt_01.log", "pid": 52896, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T09:43:58.683379+00:00", "started_monotonic": 635273.157367747, "task": "social", "unit_id": "social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:30.963830+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-mlp", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_health": {"failures": [], "format_ok": true, "max_abs": 80000.0, "nonfinite_tensor_count": 0, "ok": true, "path": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "tensor_count": 66}, "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0009_attempt_01.log", "pid": 52898, "resume_from": null, "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T09:43:58.684978+00:00", "started_monotonic": 635273.158965726, "task": "social", "unit_id": "social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-02T11:50:30.970766+00:00"} diff --git a/report.log b/report.log new file mode 100644 index 0000000000000000000000000000000000000000..6077a957c80a14e90c599cfcfa671f215e380a5c --- /dev/null +++ b/report.log @@ -0,0 +1,71 @@ +{ + "analysis_script": "/root/mouse-run-run-perf/scripts/analysis/summarize_experiment.py", + "generated_at": "2026-07-02T11:51:02.815710+00:00", + "input_hashes": { + "evaluations.jsonl": "265ed5550a98149da5384ac6cbcd551db5b121176f7046a4ca58d89531a77cc7", + "exclusions.jsonl": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "rollouts.jsonl": "140147549967098f1a78dadab8e5888eb41a37389c094f79b623d22d7f0bd74a", + "runs.jsonl": "bfab762e5dd3673a8e08fa06635ebebd07bc02783f016fc26e56fee41756e2d8" + }, + "input_tables": { + "evaluations.jsonl": "runs/tables/mouse-run-run-2-mlp/evaluations.jsonl", + "exclusions.jsonl": "runs/tables/mouse-run-run-2-mlp/exclusions.jsonl", + "rollouts.jsonl": "runs/tables/mouse-run-run-2-mlp/rollouts.jsonl", + "runs.jsonl": "runs/tables/mouse-run-run-2-mlp/runs.jsonl" + }, + "output_hashes": { + "evaluation_comparison": "f884d82cdc4bac267de1ab48120986784960a6448d8c7f4502cc3b48c4369735", + "training_curves": "f79e51dd791199e40a8057983a80cb9405bc95f186a2e5ee61eae94544f41bac" + }, + "outputs": { + "figures": [ + "runs/reports/mouse-run-run-2-mlp/figures/training_curves.png", + "runs/reports/mouse-run-run-2-mlp/figures/evaluation_comparison.png" + ], + "report": "runs/reports/mouse-run-run-2-mlp/report.md", + "summary": "runs/reports/mouse-run-run-2-mlp/summary.json" + }, + "provenance": { + "command": "scripts/analysis/summarize_experiment.py runs/tables/mouse-run-run-2-mlp", + "created_at": "2026-07-02T11:51:02.816299+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "schema_version": 2 +} diff --git a/rollout.log b/rollout.log new file mode 100644 index 0000000000000000000000000000000000000000..f52fd6d11532b023c1068d129f069c8539ae7ad7 --- /dev/null +++ b/rollout.log @@ -0,0 +1,20 @@ +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0000__attempt_01__checkpoints__latest__9e74dbbabd.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0001__attempt_01__checkpoints__latest__6aac5a2c12.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0002__attempt_01__checkpoints__latest__33f1a8c09e.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0003__attempt_01__checkpoints__latest__ddb70257e3.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0004__attempt_01__checkpoints__latest__8dc5f10a59.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0005__attempt_01__checkpoints__latest__c46a11793f.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0006__attempt_01__checkpoints__latest__c29253393a.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0007__attempt_01__checkpoints__latest__a64ce3bd0c.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0008__attempt_01__checkpoints__latest__3f8cb5f7d2.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__non_social__seed_0009__attempt_01__checkpoints__latest__e145c2cab5.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0000__attempt_01__checkpoints__latest__f2865fcba9.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0001__attempt_01__checkpoints__latest__61848b4a30.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0002__attempt_01__checkpoints__latest__77d0837868.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0003__attempt_01__checkpoints__latest__3d4e5109f0.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0004__attempt_01__checkpoints__latest__4399b2519b.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0005__attempt_01__checkpoints__latest__7ddcdb43f2.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0006__attempt_01__checkpoints__latest__b87e617200.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0007__attempt_01__checkpoints__latest__b494d062b3.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0008__attempt_01__checkpoints__latest__c502c04b75.safetensors +saved_rollouts=runs/mouse-run-run-2-mlp/paper_rollouts/mouse-run-run-2-mlp__social__seed_0009__attempt_01__checkpoints__latest__b92a4c3ad0.safetensors diff --git a/run_status.json b/run_status.json new file mode 100644 index 0000000000000000000000000000000000000000..c9feeb7293a727ff2d997d526734e398d48d8038 --- /dev/null +++ b/run_status.json @@ -0,0 +1,989 @@ +{ + "completed_jobs": 20, + "concurrency": 10, + "elapsed_seconds": 15264.532410176937, + "experiment": "mouse-run-run-2-mlp", + "failed": [], + "failed_attempts": [], + "failed_attempts_count": 0, + "failed_jobs": 0, + "interrupted_attempts": [], + "interrupted_attempts_count": 0, + "jobs": [ + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1203057210158374, + "elapsed_seconds": 7633.100595657015, + "env_steps_per_second": 10480.668896924715, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0000_attempt_01.log", + "metrics": { + "approx_kl": 0.002805363851033249, + "chaser_grad_norm": 0.188689436763525, + "chaser_new_fields": 52.92499923706055, + "chaser_partner_vision": 0.9490000605583191, + "chaser_return": 23.260000228881836, + "chaser_subspace_norm": 46.499481201171875, + "collisions_per_episode": 21.450000762939453, + "entropy": 0.1341038942337036, + "explorer_grad_norm": 1.1695344895124435, + "explorer_new_fields": 65.55000305175781, + "explorer_partner_vision": 0.9490000605583191, + "explorer_return": 31.037500381469727, + "explorer_subspace_norm": 105.87859344482422, + "final_distance": 1.655848741531372, + "max_param_abs": 1.2022045850753784, + "policy_loss": -0.003802475623324142, + "value_loss": 4.788339614868164 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", + "schema_version": 2, + "seed": 0, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T09:43:30.660267+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0000", + "unix_time": 1782985410.6602778, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.111703482349177, + "elapsed_seconds": 7602.132536457037, + "env_steps_per_second": 10523.362966424142, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0000_attempt_01.log", + "metrics": { + "approx_kl": 0.0002434901476933704, + "chaser_grad_norm": 0.07690595369786024, + "chaser_new_fields": 83.0999984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.540000915527344, + "chaser_subspace_norm": 75.278076171875, + "collisions_per_episode": 1.25, + "entropy": 0.03620169032365084, + "explorer_grad_norm": 0.2844117097556591, + "explorer_new_fields": 75.5250015258789, + "explorer_partner_vision": 0.0, + "explorer_return": 62.76250076293945, + "explorer_subspace_norm": 28.751102447509766, + "final_distance": 5.306502342224121, + "max_param_abs": 1.7384926080703735, + "policy_loss": -0.0007488939012390006, + "value_loss": 3.4187601804733276 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0000.attempt_01", + "schema_version": 2, + "seed": 0, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T09:42:59.855609+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0000", + "unix_time": 1782985379.8556223, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.123490096754459, + "elapsed_seconds": 7644.564348316053, + "env_steps_per_second": 10464.952135254434, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0001_attempt_01.log", + "metrics": { + "approx_kl": 0.0005562261293015069, + "chaser_grad_norm": 0.6423081681132317, + "chaser_new_fields": 46.20000076293945, + "chaser_partner_vision": 0.9297500252723694, + "chaser_return": 23.4325008392334, + "chaser_subspace_norm": 57.10129928588867, + "collisions_per_episode": 22.625, + "entropy": 0.14692536741495132, + "explorer_grad_norm": 0.5817638859152794, + "explorer_new_fields": 52.775001525878906, + "explorer_partner_vision": 0.9297500252723694, + "explorer_return": 12.71250057220459, + "explorer_subspace_norm": 122.46495819091797, + "final_distance": 1.791115403175354, + "max_param_abs": 1.3355721235275269, + "policy_loss": -0.004476253691127141, + "value_loss": 5.1513988971710205 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", + "schema_version": 2, + "seed": 1, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T09:43:42.321981+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0001", + "unix_time": 1782985422.3219912, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.116271049076135, + "elapsed_seconds": 7618.575776674086, + "env_steps_per_second": 10500.650298043534, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0001_attempt_01.log", + "metrics": { + "approx_kl": 0.0007710583850932262, + "chaser_grad_norm": 0.11076641827821732, + "chaser_new_fields": 83.25, + "chaser_partner_vision": 0.0, + "chaser_return": 6.595000743865967, + "chaser_subspace_norm": 107.82763671875, + "collisions_per_episode": 0.9750000238418579, + "entropy": 0.06922571919858456, + "explorer_grad_norm": 0.3988708183169365, + "explorer_new_fields": 73.5, + "explorer_partner_vision": 0.0, + "explorer_return": 59.76250076293945, + "explorer_subspace_norm": 27.074764251708984, + "final_distance": 4.443612575531006, + "max_param_abs": 1.3260666131973267, + "policy_loss": -0.0006433645908252572, + "value_loss": 3.4030256271362305 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0001.attempt_01", + "schema_version": 2, + "seed": 1, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T09:43:16.348219+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0001", + "unix_time": 1782985396.3482296, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.122357136064416, + "elapsed_seconds": 7640.485689831898, + "env_steps_per_second": 10470.538555744632, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0002_attempt_01.log", + "metrics": { + "approx_kl": 0.0014266868641836794, + "chaser_grad_norm": 0.20290611311793327, + "chaser_new_fields": 49.29999923706055, + "chaser_partner_vision": 0.9287500381469727, + "chaser_return": 18.48750114440918, + "chaser_subspace_norm": 48.7114143371582, + "collisions_per_episode": 17.625, + "entropy": 0.18204283714294434, + "explorer_grad_norm": 0.6118505597114563, + "explorer_new_fields": 60.95000076293945, + "explorer_partner_vision": 0.9287500381469727, + "explorer_return": 26.837499618530273, + "explorer_subspace_norm": 104.41957092285156, + "final_distance": 1.6572967767715454, + "max_param_abs": 1.1167881488800049, + "policy_loss": -0.005259661337355226, + "value_loss": 5.285007953643799 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", + "schema_version": 2, + "seed": 2, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T09:43:38.236370+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0002", + "unix_time": 1782985418.2363846, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1173361086988653, + "elapsed_seconds": 7622.409991315915, + "env_steps_per_second": 10495.368274750725, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0002_attempt_01.log", + "metrics": { + "approx_kl": 0.00015983218477721994, + "chaser_grad_norm": 0.06889200676232576, + "chaser_new_fields": 84.3499984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.7600016593933105, + "chaser_subspace_norm": 98.9588623046875, + "collisions_per_episode": 2.075000047683716, + "entropy": 0.02094564540311694, + "explorer_grad_norm": 1.1887003779411316, + "explorer_new_fields": 82.75, + "explorer_partner_vision": 0.0, + "explorer_return": 73.30000305175781, + "explorer_subspace_norm": 38.70703887939453, + "final_distance": 4.667576789855957, + "max_param_abs": 1.5199339389801025, + "policy_loss": -0.0006238509232616707, + "value_loss": 3.509905695915222 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0002.attempt_01", + "schema_version": 2, + "seed": 2, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T09:43:20.029624+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0002", + "unix_time": 1782985400.0296402, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1228015855661297, + "elapsed_seconds": 7642.0857080380665, + "env_steps_per_second": 10468.3463463194, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0003_attempt_01.log", + "metrics": { + "approx_kl": 0.000801136315581541, + "chaser_grad_norm": 0.38031933456659317, + "chaser_new_fields": 52.5, + "chaser_partner_vision": 0.9297500252723694, + "chaser_return": 21.880001068115234, + "chaser_subspace_norm": 52.41925811767578, + "collisions_per_episode": 20.200000762939453, + "entropy": 0.12773804366588593, + "explorer_grad_norm": 0.23243309557437897, + "explorer_new_fields": 59.29999923706055, + "explorer_partner_vision": 0.9297500252723694, + "explorer_return": 22.887500762939453, + "explorer_subspace_norm": 119.99281311035156, + "final_distance": 1.4895528554916382, + "max_param_abs": 1.4659936428070068, + "policy_loss": -0.0027591757117768623, + "value_loss": 5.230631113052368 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", + "schema_version": 2, + "seed": 3, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T09:43:40.053787+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0003", + "unix_time": 1782985420.0537992, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.119901634450265, + "elapsed_seconds": 7631.6458840209525, + "env_steps_per_second": 10482.666677119156, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0003_attempt_01.log", + "metrics": { + "approx_kl": 0.0002841827333717878, + "chaser_grad_norm": 0.04017048981040716, + "chaser_new_fields": 87.07500457763672, + "chaser_partner_vision": 0.0, + "chaser_return": 7.300001621246338, + "chaser_subspace_norm": 112.96394348144531, + "collisions_per_episode": 1.725000023841858, + "entropy": 0.024153232108801603, + "explorer_grad_norm": 0.8466866165399551, + "explorer_new_fields": 75.2249984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 62.125, + "explorer_subspace_norm": 36.111541748046875, + "final_distance": 4.144174098968506, + "max_param_abs": 1.6294050216674805, + "policy_loss": -0.0005663236785924397, + "value_loss": 3.4905534386634827 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0003.attempt_01", + "schema_version": 2, + "seed": 3, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T09:43:29.377490+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0003", + "unix_time": 1782985409.377503, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.122254997810556, + "elapsed_seconds": 7640.117992118001, + "env_steps_per_second": 10471.042473759273, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0004_attempt_01.log", + "metrics": { + "approx_kl": 0.000236154562574864, + "chaser_grad_norm": 0.38722849637269974, + "chaser_new_fields": 45.57500076293945, + "chaser_partner_vision": 0.9225000143051147, + "chaser_return": 19.41750144958496, + "chaser_subspace_norm": 58.249393463134766, + "collisions_per_episode": 19.125, + "entropy": 0.19220001995563507, + "explorer_grad_norm": 0.3825383335351944, + "explorer_new_fields": 56.85000228881836, + "explorer_partner_vision": 0.9225000143051147, + "explorer_return": 19.30000114440918, + "explorer_subspace_norm": 90.54417419433594, + "final_distance": 1.830224633216858, + "max_param_abs": 1.0662928819656372, + "policy_loss": -0.004796283046713867, + "value_loss": 5.130034923553467 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", + "schema_version": 2, + "seed": 4, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T09:43:38.099176+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0004", + "unix_time": 1782985418.0991888, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.122494922226114, + "elapsed_seconds": 7640.98172001401, + "env_steps_per_second": 10469.858839009672, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0004_attempt_01.log", + "metrics": { + "approx_kl": 0.00037036755928841064, + "chaser_grad_norm": 0.08058446925133467, + "chaser_new_fields": 82.17500305175781, + "chaser_partner_vision": 0.0, + "chaser_return": 6.330001354217529, + "chaser_subspace_norm": 92.44800567626953, + "collisions_per_episode": 1.8250000476837158, + "entropy": 0.04575740918517113, + "explorer_grad_norm": 0.3279200941324234, + "explorer_new_fields": 77.9749984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 66.0999984741211, + "explorer_subspace_norm": 29.62271499633789, + "final_distance": 5.308696269989014, + "max_param_abs": 1.3726928234100342, + "policy_loss": -0.0006299224167634065, + "value_loss": 3.322889506816864 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0004.attempt_01", + "schema_version": 2, + "seed": 4, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T09:43:38.632658+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0004", + "unix_time": 1782985418.632673, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.0991321679530666, + "elapsed_seconds": 7556.8758046310395, + "env_steps_per_second": 10586.385441318756, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0005_attempt_01.log", + "metrics": { + "approx_kl": 0.0005827252787078407, + "chaser_grad_norm": 0.3728112056851387, + "chaser_new_fields": 51.47500228881836, + "chaser_partner_vision": 0.95250004529953, + "chaser_return": 20.117502212524414, + "chaser_subspace_norm": 50.84516143798828, + "collisions_per_episode": 18.575000762939453, + "entropy": 0.13566195219755173, + "explorer_grad_norm": 0.6062372475862503, + "explorer_new_fields": 58.57500076293945, + "explorer_partner_vision": 0.95250004529953, + "explorer_return": 20.774999618530273, + "explorer_subspace_norm": 104.22787475585938, + "final_distance": 1.5808099508285522, + "max_param_abs": 1.3590010404586792, + "policy_loss": -0.003127762281341262, + "value_loss": 4.4600828886032104 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", + "schema_version": 2, + "seed": 5, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T11:49:16.134110+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0005", + "unix_time": 1782992956.134124, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1006427376291765, + "elapsed_seconds": 7562.313855465036, + "env_steps_per_second": 10578.772784230137, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0005_attempt_01.log", + "metrics": { + "approx_kl": 0.00022429173202174373, + "chaser_grad_norm": 0.04249562695622444, + "chaser_new_fields": 80.4749984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.015000820159912, + "chaser_subspace_norm": 93.73216247558594, + "collisions_per_episode": 1.024999976158142, + "entropy": 0.05529557354748249, + "explorer_grad_norm": 0.39152009785175323, + "explorer_new_fields": 73.2249984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 59.42499923706055, + "explorer_subspace_norm": 117.07512664794922, + "final_distance": 4.730400562286377, + "max_param_abs": 1.606895923614502, + "policy_loss": -0.0002411644566997495, + "value_loss": 3.523293375968933 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0005.attempt_01", + "schema_version": 2, + "seed": 5, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T11:49:31.723491+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0005", + "unix_time": 1782992971.7235026, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.099672739202215, + "elapsed_seconds": 7558.8218611279735, + "env_steps_per_second": 10583.659923434405, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0006_attempt_01.log", + "metrics": { + "approx_kl": 0.00025015939701367174, + "chaser_grad_norm": 0.31481391191482544, + "chaser_new_fields": 49.900001525878906, + "chaser_partner_vision": 0.9252500534057617, + "chaser_return": 21.020002365112305, + "chaser_subspace_norm": 49.21955108642578, + "collisions_per_episode": 19.80000114440918, + "entropy": 0.1560382954776287, + "explorer_grad_norm": 0.6951030343770981, + "explorer_new_fields": 59.42499923706055, + "explorer_partner_vision": 0.9252500534057617, + "explorer_return": 22.48750114440918, + "explorer_subspace_norm": 113.72798919677734, + "final_distance": 1.6350990533828735, + "max_param_abs": 1.6354095935821533, + "policy_loss": -0.0035889741121788443, + "value_loss": 4.686129331588745 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", + "schema_version": 2, + "seed": 6, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T11:49:38.255928+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0006", + "unix_time": 1782992978.2559416, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1019998172186187, + "elapsed_seconds": 7567.199341987027, + "env_steps_per_second": 10571.942985050697, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0006_attempt_01.log", + "metrics": { + "approx_kl": 8.635803430578903e-06, + "chaser_grad_norm": 0.10996681079268456, + "chaser_new_fields": 82.375, + "chaser_partner_vision": 0.0, + "chaser_return": 6.390000820159912, + "chaser_subspace_norm": 73.99776458740234, + "collisions_per_episode": 1.149999976158142, + "entropy": 0.05159799661487341, + "explorer_grad_norm": 0.7565127648413181, + "explorer_new_fields": 61.025001525878906, + "explorer_partner_vision": 0.0, + "explorer_return": 41.23749923706055, + "explorer_subspace_norm": 46.546077728271484, + "final_distance": 4.668344020843506, + "max_param_abs": 1.2401388883590698, + "policy_loss": -0.0002673804239634592, + "value_loss": 3.8410890698432922 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0006.attempt_01", + "schema_version": 2, + "seed": 6, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T11:49:56.653691+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0006", + "unix_time": 1782992996.6537013, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.10750246579028, + "elapsed_seconds": 7587.008876845008, + "env_steps_per_second": 10544.33984440879, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0007_attempt_01.log", + "metrics": { + "approx_kl": 0.00013653355120357702, + "chaser_grad_norm": 0.9264960587024689, + "chaser_new_fields": 46.92499923706055, + "chaser_partner_vision": 0.9470000267028809, + "chaser_return": 24.285001754760742, + "chaser_subspace_norm": 67.47857666015625, + "collisions_per_episode": 23.200000762939453, + "entropy": 0.1407805271446705, + "explorer_grad_norm": 0.5267750471830368, + "explorer_new_fields": 55.900001525878906, + "explorer_partner_vision": 0.9470000267028809, + "explorer_return": 17.149999618530273, + "explorer_subspace_norm": 89.71617889404297, + "final_distance": 1.689988374710083, + "max_param_abs": 1.2198518514633179, + "policy_loss": -0.003363337086513596, + "value_loss": 4.716352343559265 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", + "schema_version": 2, + "seed": 7, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T11:50:16.462261+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0007", + "unix_time": 1782993016.4622712, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1043706129752615, + "elapsed_seconds": 7575.734206710942, + "env_steps_per_second": 10560.032574681967, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0007_attempt_01.log", + "metrics": { + "approx_kl": 0.0005693047922270544, + "chaser_grad_norm": 0.08362068422138691, + "chaser_new_fields": 82.0250015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 6.3000006675720215, + "chaser_subspace_norm": 109.71160888671875, + "collisions_per_episode": 2.174999952316284, + "entropy": 0.05865261983126402, + "explorer_grad_norm": 0.6362885981798172, + "explorer_new_fields": 60.57500076293945, + "explorer_partner_vision": 0.0, + "explorer_return": 40.07500076293945, + "explorer_subspace_norm": 35.61513900756836, + "final_distance": 4.870059967041016, + "max_param_abs": 1.4202895164489746, + "policy_loss": -0.0008471746835780891, + "value_loss": 4.014994144439697 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0007.attempt_01", + "schema_version": 2, + "seed": 7, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T11:50:15.587272+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0007", + "unix_time": 1782993015.5872822, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.1067731915308587, + "elapsed_seconds": 7584.383489511092, + "env_steps_per_second": 10547.989841314971, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0008_attempt_01.log", + "metrics": { + "approx_kl": 0.0022941239069333985, + "chaser_grad_norm": 0.43133843690156937, + "chaser_new_fields": 50.125, + "chaser_partner_vision": 0.9205000400543213, + "chaser_return": 20.852500915527344, + "chaser_subspace_norm": 52.702186584472656, + "collisions_per_episode": 19.67500114440918, + "entropy": 0.15483620017766953, + "explorer_grad_norm": 0.3490636646747589, + "explorer_new_fields": 60.525001525878906, + "explorer_partner_vision": 0.9205000400543213, + "explorer_return": 23.0, + "explorer_subspace_norm": 101.96681213378906, + "final_distance": 1.8305482864379883, + "max_param_abs": 1.1802449226379395, + "policy_loss": -0.00493374926503698, + "value_loss": 4.697651386260986 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", + "schema_version": 2, + "seed": 8, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T11:50:24.136096+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0008", + "unix_time": 1782993024.136106, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.10492892708251, + "elapsed_seconds": 7577.744137497037, + "env_steps_per_second": 10557.231617802336, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0008_attempt_01.log", + "metrics": { + "approx_kl": 3.927184380664306e-06, + "chaser_grad_norm": 0.07202051393687725, + "chaser_new_fields": 81.3499984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.145000457763672, + "chaser_subspace_norm": 132.7053680419922, + "collisions_per_episode": 1.975000023841858, + "entropy": 0.13418130949139595, + "explorer_grad_norm": 0.6797178164124489, + "explorer_new_fields": 68.7249984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 52.48749923706055, + "explorer_subspace_norm": 61.86991500854492, + "final_distance": 4.521096229553223, + "max_param_abs": 1.4186153411865234, + "policy_loss": -0.0009259324722812323, + "value_loss": 4.002953112125397 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0008.attempt_01", + "schema_version": 2, + "seed": 8, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T11:50:17.380387+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0008", + "unix_time": 1782993017.380397, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.107122335001089, + "elapsed_seconds": 7585.64040600392, + "env_steps_per_second": 10546.242072941028, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/social_seed_0009_attempt_01.log", + "metrics": { + "approx_kl": -0.0002899035942244277, + "chaser_grad_norm": 0.2534610331058502, + "chaser_new_fields": 51.625, + "chaser_partner_vision": 0.9552500247955322, + "chaser_return": 22.76500129699707, + "chaser_subspace_norm": 57.69636535644531, + "collisions_per_episode": 21.25, + "entropy": 0.1398237645626068, + "explorer_grad_norm": 0.3390047997236252, + "explorer_new_fields": 58.42499923706055, + "explorer_partner_vision": 0.9552500247955322, + "explorer_return": 21.6875, + "explorer_subspace_norm": 105.33492279052734, + "final_distance": 1.5442389249801636, + "max_param_abs": 1.1976157426834106, + "policy_loss": -0.002912267684005876, + "value_loss": 4.631349444389343 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", + "schema_version": 2, + "seed": 9, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T11:50:25.494653+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0009", + "unix_time": 1782993025.494663, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 2.102990282865293, + "elapsed_seconds": 7570.7650183150545, + "env_steps_per_second": 10566.96381494677, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-mlp/logs/non_social_seed_0009_attempt_01.log", + "metrics": { + "approx_kl": 0.00014854547405664803, + "chaser_grad_norm": 0.09859942272305489, + "chaser_new_fields": 82.5999984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 6.340001106262207, + "chaser_subspace_norm": 98.81183624267578, + "collisions_per_episode": 2.7250001430511475, + "entropy": 0.07547229155898094, + "explorer_grad_norm": 0.32836734503507614, + "explorer_new_fields": 65.67500305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 47.462501525878906, + "explorer_subspace_norm": 26.826025009155273, + "final_distance": 5.683727741241455, + "max_param_abs": 1.2476426362991333, + "policy_loss": -0.0011177757633866747, + "value_loss": 3.991724371910095 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-mlp/non_social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-mlp.non_social.seed_0009.attempt_01", + "schema_version": 2, + "seed": 9, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T11:50:10.506360+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0009", + "unix_time": 1782993010.5063717, + "update": 20000 + } + ], + "known_attempts": 20, + "observed_updates": 400000, + "pending_jobs": 0, + "progress_fraction": 1.0, + "running": [], + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:40.978924+00:00", + "total_jobs": 20, + "total_updates": 400000, + "unix_time": 1782993040.9789648, + "valid_pairs_by_task": { + "non_social": 10, + "social": 10 + } +} diff --git a/runner.log b/runner.log new file mode 100644 index 0000000000000000000000000000000000000000..03d79ad257ba4f7045515d3528c1209404c07013 --- /dev/null +++ b/runner.log @@ -0,0 +1,40 @@ +started task=social seed=0 attempt=1 pid=52153 log=runs/mouse-run-run-2-mlp/logs/social_seed_0000_attempt_01.log +started task=non_social seed=0 attempt=1 pid=52154 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0000_attempt_01.log +started task=social seed=1 attempt=1 pid=52155 log=runs/mouse-run-run-2-mlp/logs/social_seed_0001_attempt_01.log +started task=non_social seed=1 attempt=1 pid=52156 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0001_attempt_01.log +started task=social seed=2 attempt=1 pid=52157 log=runs/mouse-run-run-2-mlp/logs/social_seed_0002_attempt_01.log +started task=non_social seed=2 attempt=1 pid=52158 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0002_attempt_01.log +started task=social seed=3 attempt=1 pid=52159 log=runs/mouse-run-run-2-mlp/logs/social_seed_0003_attempt_01.log +started task=non_social seed=3 attempt=1 pid=52160 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0003_attempt_01.log +started task=social seed=4 attempt=1 pid=52161 log=runs/mouse-run-run-2-mlp/logs/social_seed_0004_attempt_01.log +started task=non_social seed=4 attempt=1 pid=52162 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0004_attempt_01.log +completed task=non_social seed=0 attempt=1 +started task=social seed=5 attempt=1 pid=52725 log=runs/mouse-run-run-2-mlp/logs/social_seed_0005_attempt_01.log +completed task=non_social seed=1 attempt=1 +started task=non_social seed=5 attempt=1 pid=52759 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0005_attempt_01.log +completed task=non_social seed=2 attempt=1 +started task=social seed=6 attempt=1 pid=52793 log=runs/mouse-run-run-2-mlp/logs/social_seed_0006_attempt_01.log +completed task=social seed=0 attempt=1 +completed task=non_social seed=3 attempt=1 +started task=non_social seed=6 attempt=1 pid=52827 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0006_attempt_01.log +started task=social seed=7 attempt=1 pid=52828 log=runs/mouse-run-run-2-mlp/logs/social_seed_0007_attempt_01.log +completed task=social seed=1 attempt=1 +completed task=social seed=2 attempt=1 +completed task=social seed=3 attempt=1 +completed task=social seed=4 attempt=1 +completed task=non_social seed=4 attempt=1 +started task=non_social seed=7 attempt=1 pid=52895 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0007_attempt_01.log +started task=social seed=8 attempt=1 pid=52896 log=runs/mouse-run-run-2-mlp/logs/social_seed_0008_attempt_01.log +started task=non_social seed=8 attempt=1 pid=52897 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0008_attempt_01.log +started task=social seed=9 attempt=1 pid=52898 log=runs/mouse-run-run-2-mlp/logs/social_seed_0009_attempt_01.log +started task=non_social seed=9 attempt=1 pid=52899 log=runs/mouse-run-run-2-mlp/logs/non_social_seed_0009_attempt_01.log +completed task=social seed=5 attempt=1 +completed task=non_social seed=5 attempt=1 +completed task=social seed=6 attempt=1 +completed task=non_social seed=6 attempt=1 +completed task=social seed=7 attempt=1 +completed task=non_social seed=7 attempt=1 +completed task=non_social seed=8 attempt=1 +completed task=non_social seed=9 attempt=1 +completed task=social seed=8 attempt=1 +completed task=social seed=9 attempt=1 diff --git a/social/seed_0000/attempt_01/checkpoints/latest.safetensors b/social/seed_0000/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fe5a48d8f26b1bb10717828b8b5a1fded9f2f44b --- /dev/null +++ b/social/seed_0000/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f4bcf1374937905b5d1f57c73129408305c6289a4aa0a054328102a1a4ceded0 +size 11460408 diff --git a/social/seed_0000/attempt_01/config.json b/social/seed_0000/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..1972a822e9ebbaad2bdacf890072db7596ef9163 --- /dev/null +++ b/social/seed_0000/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0000.attempt_01", + "seed": 0, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0000/attempt_01/metrics.jsonl b/social/seed_0000/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6cf6cf670fe4f7bd63ea51058ffe7a4b6c0d7486 --- /dev/null +++ b/social/seed_0000/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.00040886967779240675, "elapsed_seconds": 1.4719308400526643, "env_steps_per_second": 2717.5189833354425, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 8.176984686170343, "eta_seconds": 29437.144870213233, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -7.231411298547208e-06, "chaser_grad_norm": 0.17073261365294456, "chaser_new_fields": 35.875, "chaser_partner_vision": 0.3425000309944153, "chaser_return": -1.1800003051757812, "chaser_subspace_norm": 0.0034334014635533094, "collisions_per_episode": 1.5500000715255737, "entropy": 2.7725025415420532, "explorer_grad_norm": 0.14562124386429787, "explorer_new_fields": 33.875, "explorer_partner_vision": 0.3425000309944153, "explorer_return": -0.4124999940395355, "explorer_subspace_norm": 0.0028507569804787636, "final_distance": 5.3778157234191895, "max_param_abs": 0.15428674221038818, "policy_loss": -0.0007909450682346275, "value_loss": 3.1282492876052856}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.031606+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.0316236, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05228271625722603, "elapsed_seconds": 188.21777852601372, "env_steps_per_second": 10625.988765049517, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.039025934031815, "eta_seconds": 7340.493362514535, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0007004108923813757, "chaser_grad_norm": 0.5508566275238991, "chaser_new_fields": 54.47500228881836, "chaser_partner_vision": 0.8752500414848328, "chaser_return": 15.032501220703125, "chaser_subspace_norm": 16.14761734008789, "collisions_per_episode": 13.375, "entropy": 0.6327623873949051, "explorer_grad_norm": 0.2901584357023239, "explorer_new_fields": 64.42500305175781, "explorer_partner_vision": 0.8752500414848328, "explorer_return": 35.67499923706055, "explorer_subspace_norm": 8.446002006530762, "final_distance": 2.0203511714935303, "max_param_abs": 0.3101498782634735, "policy_loss": -0.006472172427933742, "value_loss": 5.703834056854248}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:25.777450+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977965.7774615, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10450647935054601, "elapsed_seconds": 376.22332566196565, "env_steps_per_second": 10631.98299297895, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9856231076603743, "eta_seconds": 7148.243187577347, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0011974972750881108, "chaser_grad_norm": 0.4391440153121948, "chaser_new_fields": 52.70000076293945, "chaser_partner_vision": 0.8675000667572021, "chaser_return": 14.472500801086426, "chaser_subspace_norm": 22.766597747802734, "collisions_per_episode": 13.225000381469727, "entropy": 0.4586358517408371, "explorer_grad_norm": 0.43156542629003525, "explorer_new_fields": 63.275001525878906, "explorer_partner_vision": 0.8675000667572021, "explorer_return": 33.01250076293945, "explorer_subspace_norm": 13.137873649597168, "final_distance": 2.4646008014678955, "max_param_abs": 0.3303264081478119, "policy_loss": -0.007705260874701025, "value_loss": 5.705733418464661}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:33.782999+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978153.7830143, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15716971717305972, "elapsed_seconds": 565.810981823015, "env_steps_per_second": 10604.248048824178, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9384265118010702, "eta_seconds": 6978.335442483853, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0006259187082893192, "chaser_grad_norm": 0.25775307416915894, "chaser_new_fields": 52.625, "chaser_partner_vision": 0.8442500233650208, "chaser_return": 15.30250072479248, "chaser_subspace_norm": 25.7149600982666, "collisions_per_episode": 14.074999809265137, "entropy": 0.37239792943000793, "explorer_grad_norm": 0.2852398492395878, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.8442500233650208, "explorer_return": 35.13750076293945, "explorer_subspace_norm": 17.414716720581055, "final_distance": 2.225663423538208, "max_param_abs": 0.36135244369506836, "policy_loss": -0.006344570733606503, "value_loss": 5.844202876091003}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:43.370653+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978343.3706632, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.2099371798611052, "elapsed_seconds": 755.7738474999787, "env_steps_per_second": 10585.17706912348, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.889434618749947, "eta_seconds": 6801.964627499809, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0011572397580028593, "chaser_grad_norm": 0.4071383848786354, "chaser_new_fields": 49.57500076293945, "chaser_partner_vision": 0.9332500696182251, "chaser_return": 16.170001983642578, "chaser_subspace_norm": 30.913875579833984, "collisions_per_episode": 15.449999809265137, "entropy": 0.28281812369823456, "explorer_grad_norm": 0.45905886590480804, "explorer_new_fields": 62.400001525878906, "explorer_partner_vision": 0.9332500696182251, "explorer_return": 29.42500114440918, "explorer_subspace_norm": 24.365882873535156, "final_distance": 2.1872448921203613, "max_param_abs": 0.41835716366767883, "policy_loss": -0.005352281774857914, "value_loss": 5.605498909950256}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:48:53.333519+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978533.3335345, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26319286538029296, "elapsed_seconds": 947.4943153690547, "env_steps_per_second": 10554.153030569836, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8423500576620508, "eta_seconds": 6632.460207583383, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0010914268331818633, "chaser_grad_norm": 0.4338730685412884, "chaser_new_fields": 50.47500228881836, "chaser_partner_vision": 0.924250066280365, "chaser_return": 17.930002212524414, "chaser_subspace_norm": 31.0838623046875, "collisions_per_episode": 16.850000381469727, "entropy": 0.24761737883090973, "explorer_grad_norm": 0.22862951084971428, "explorer_new_fields": 60.900001525878906, "explorer_partner_vision": 0.924250066280365, "explorer_return": 26.55000114440918, "explorer_subspace_norm": 28.563730239868164, "final_distance": 1.9120800495147705, "max_param_abs": 0.41357728838920593, "policy_loss": -0.004575131649616537, "value_loss": 5.638877272605896}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:05.053986+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978725.0539978, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3152505194980444, "elapsed_seconds": 1134.90187019296, "env_steps_per_second": 10573.601396885282, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7864196104889185, "eta_seconds": 6431.110597760106, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005922479628575061, "chaser_grad_norm": 0.408519871532917, "chaser_new_fields": 49.54999923706055, "chaser_partner_vision": 0.9342500567436218, "chaser_return": 16.337499618530273, "chaser_subspace_norm": 33.83238220214844, "collisions_per_episode": 15.625, "entropy": 0.19292935356497765, "explorer_grad_norm": 0.3779001906514168, "explorer_new_fields": 63.875, "explorer_partner_vision": 0.9342500567436218, "explorer_return": 31.662500381469727, "explorer_subspace_norm": 32.78867721557617, "final_distance": 1.9254854917526245, "max_param_abs": 0.44604021310806274, "policy_loss": -0.004625527589579903, "value_loss": 5.335520267486572}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:12.461541+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978912.4615567, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3686635666791699, "elapsed_seconds": 1327.1888400450116, "env_steps_per_second": 10548.611906294504, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7379853857732295, "eta_seconds": 6256.747388783626, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.001515843262221539, "chaser_grad_norm": 0.726634293794632, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.940000057220459, "chaser_return": 21.4950008392334, "chaser_subspace_norm": 33.90365219116211, "collisions_per_episode": 20.5, "entropy": 0.20392829179763794, "explorer_grad_norm": 0.4465930536389351, "explorer_new_fields": 62.20000076293945, "explorer_partner_vision": 0.940000057220459, "explorer_return": 25.25, "explorer_subspace_norm": 35.112789154052734, "final_distance": 1.8083772659301758, "max_param_abs": 0.5109430551528931, "policy_loss": -0.005126612390391472, "value_loss": 5.556533098220825}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:24.748510+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979104.7485216, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4225727539883357, "elapsed_seconds": 1521.2619143580087, "env_steps_per_second": 10517.584019548796, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6902910159533426, "eta_seconds": 6085.047657432034, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.000952489059478534, "chaser_grad_norm": 0.5684905275702477, "chaser_new_fields": 50.57500076293945, "chaser_partner_vision": 0.9540000557899475, "chaser_return": 22.48000144958496, "chaser_subspace_norm": 34.81595993041992, "collisions_per_episode": 21.05000114440918, "entropy": 0.17878960072994232, "explorer_grad_norm": 0.38309396058321, "explorer_new_fields": 59.95000076293945, "explorer_partner_vision": 0.9540000557899475, "explorer_return": 22.3125, "explorer_subspace_norm": 39.81028747558594, "final_distance": 1.6400043964385986, "max_param_abs": 0.520993173122406, "policy_loss": -0.004937307518908085, "value_loss": 5.912084102630615}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:38.821587+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979298.8216016, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4770561657055643, "elapsed_seconds": 1717.4021965400316, "env_steps_per_second": 10480.946185036762, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6431934596524993, "eta_seconds": 5915.496454748997, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0007124856197637808, "chaser_grad_norm": 0.4921956919133663, "chaser_new_fields": 48.04999923706055, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 17.327499389648438, "chaser_subspace_norm": 31.86878776550293, "collisions_per_episode": 16.825000762939453, "entropy": 0.22870143502950668, "explorer_grad_norm": 0.40463902801275253, "explorer_new_fields": 62.92499923706055, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 28.962499618530273, "explorer_subspace_norm": 41.776268005371094, "final_distance": 1.8810726404190063, "max_param_abs": 0.5208597779273987, "policy_loss": -0.005653111043383063, "value_loss": 5.590240955352783}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:04:54.961869+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979494.9618819, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.530341903659735, "elapsed_seconds": 1909.230853175046, "env_steps_per_second": 10475.42258535161, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.591025710979205, "eta_seconds": 5727.692559525138, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0003629874691091217, "chaser_grad_norm": 0.42813603207468987, "chaser_new_fields": 50.07500076293945, "chaser_partner_vision": 0.9305000305175781, "chaser_return": 21.207502365112305, "chaser_subspace_norm": 31.802772521972656, "collisions_per_episode": 20.125, "entropy": 0.1593104749917984, "explorer_grad_norm": 0.34928206354379654, "explorer_new_fields": 63.92499923706055, "explorer_partner_vision": 0.9305000305175781, "explorer_return": 28.25, "explorer_subspace_norm": 44.51356506347656, "final_distance": 1.6856319904327393, "max_param_abs": 0.6046883463859558, "policy_loss": -0.003571252346972642, "value_loss": 5.518453001976013}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:06.790524+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979686.7905393, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5849268272447322, "elapsed_seconds": 2105.736578081036, "env_steps_per_second": 10447.650588873117, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5420798172815668, "eta_seconds": 5551.487342213641, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0001910958976103494, "chaser_grad_norm": 0.7319878414273262, "chaser_new_fields": 47.375, "chaser_partner_vision": 0.9295000433921814, "chaser_return": 16.46750259399414, "chaser_subspace_norm": 32.06983947753906, "collisions_per_episode": 16.225000381469727, "entropy": 0.17348982766270638, "explorer_grad_norm": 1.2841002941131592, "explorer_new_fields": 57.275001525878906, "explorer_partner_vision": 0.9295000433921814, "explorer_return": 20.86250114440918, "explorer_subspace_norm": 49.87643051147461, "final_distance": 1.6583225727081299, "max_param_abs": 0.6597219705581665, "policy_loss": -0.0036080360212409346, "value_loss": 5.826892256736755}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:23.296249+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979883.2962606, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6377824455736036, "elapsed_seconds": 2296.0168040649733, "env_steps_per_second": 10452.885169441835, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.488159039671742, "eta_seconds": 5357.372542818271, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006426259686026503, "chaser_grad_norm": 0.1840935219079256, "chaser_new_fields": 48.79999923706055, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 21.587499618530273, "chaser_subspace_norm": 31.985267639160156, "collisions_per_episode": 20.875, "entropy": 0.1408773809671402, "explorer_grad_norm": 0.3943624943494797, "explorer_new_fields": 62.35000228881836, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 23.86250114440918, "explorer_subspace_norm": 47.225467681884766, "final_distance": 1.518842101097107, "max_param_abs": 0.6968187689781189, "policy_loss": -0.0030413637508139857, "value_loss": 5.077256679534912}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:33.576476+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980073.5764902, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6916645797147389, "elapsed_seconds": 2489.9924869730603, "env_steps_per_second": 10441.798574102002, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4365341270998424, "eta_seconds": 5171.522857559433, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006949275878522787, "chaser_grad_norm": 0.5277760475873947, "chaser_new_fields": 50.32500076293945, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 19.722501754760742, "chaser_subspace_norm": 31.1393985748291, "collisions_per_episode": 18.625, "entropy": 0.15545115619897842, "explorer_grad_norm": 0.5043764188885689, "explorer_new_fields": 63.375, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 27.162500381469727, "explorer_subspace_norm": 48.98280334472656, "final_distance": 1.6371082067489624, "max_param_abs": 0.7614264488220215, "policy_loss": -0.0034072986887707657, "value_loss": 5.26273238658905}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:47.552159+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980267.5521722, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.745271905689442, "elapsed_seconds": 2682.9788604819914, "env_steps_per_second": 10436.161243167551, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3840763962803924, "eta_seconds": 4982.675026609412, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00031431327727204916, "chaser_grad_norm": 0.4149128347635269, "chaser_new_fields": 47.79999923706055, "chaser_partner_vision": 0.9322500228881836, "chaser_return": 18.997501373291016, "chaser_subspace_norm": 30.417831420898438, "collisions_per_episode": 18.42500114440918, "entropy": 0.2204911969602108, "explorer_grad_norm": 0.8512056022882462, "explorer_new_fields": 61.67499923706055, "explorer_partner_vision": 0.9322500228881836, "explorer_return": 26.475000381469727, "explorer_subspace_norm": 50.792205810546875, "final_distance": 1.7728321552276611, "max_param_abs": 0.8148413896560669, "policy_loss": -0.005644801740537275, "value_loss": 5.619314670562744}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:21:00.538532+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980460.5385468, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7972730030711197, "elapsed_seconds": 2870.182811056031, "env_steps_per_second": 10452.295890156924, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3287883384518662, "eta_seconds": 4783.638018426718, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -0.00015276039078315762, "chaser_grad_norm": 0.5281955599784851, "chaser_new_fields": 48.35000228881836, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 22.36250114440918, "chaser_subspace_norm": 32.276790618896484, "collisions_per_episode": 21.524999618530273, "entropy": 0.17730383947491646, "explorer_grad_norm": 0.28483084589242935, "explorer_new_fields": 63.32500076293945, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 25.5625, "explorer_subspace_norm": 49.61220169067383, "final_distance": 1.655848741531372, "max_param_abs": 0.7907066345214844, "policy_loss": -0.003999513835648116, "value_loss": 5.633957505226135}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:07.742482+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980647.7424922, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8516779984039021, "elapsed_seconds": 3066.040794254048, "env_steps_per_second": 10436.912665992573, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2775169976058534, "eta_seconds": 4599.061191381072, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00045430049249262083, "chaser_grad_norm": 0.4265521466732025, "chaser_new_fields": 48.400001525878906, "chaser_partner_vision": 0.9465000629425049, "chaser_return": 18.377500534057617, "chaser_subspace_norm": 31.244808197021484, "collisions_per_episode": 17.774999618530273, "entropy": 0.14467697218060493, "explorer_grad_norm": 0.774466022849083, "explorer_new_fields": 67.32500457763672, "explorer_partner_vision": 0.9465000629425049, "explorer_return": 32.91250228881836, "explorer_subspace_norm": 49.75893020629883, "final_distance": 1.7137367725372314, "max_param_abs": 0.8096539974212646, "policy_loss": -0.0029702944862677327, "value_loss": 5.241043210029602}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:23.600466+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980843.600479, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9035520749758386, "elapsed_seconds": 3252.787469913019, "env_steps_per_second": 10452.573466445743, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2224528073202523, "eta_seconds": 4400.830106352908, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005076655700875232, "chaser_grad_norm": 0.42522676289081573, "chaser_new_fields": 46.375, "chaser_partner_vision": 0.9435000419616699, "chaser_return": 22.347501754760742, "chaser_subspace_norm": 31.059093475341797, "collisions_per_episode": 21.774999618530273, "entropy": 0.19601650536060333, "explorer_grad_norm": 0.7174044027924538, "explorer_new_fields": 60.775001525878906, "explorer_partner_vision": 0.9435000419616699, "explorer_return": 22.137500762939453, "explorer_subspace_norm": 46.66922378540039, "final_distance": 1.5514243841171265, "max_param_abs": 0.7923919558525085, "policy_loss": -0.003933542573360427, "value_loss": 5.629079699516296}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:30.347140+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981030.3471522, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9553852258972215, "elapsed_seconds": 3439.3868132299976, "env_steps_per_second": 10466.98203921753, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1676930538743817, "eta_seconds": 4203.6949939477745, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0004755670480491858, "chaser_grad_norm": 0.39725227653980255, "chaser_new_fields": 49.92499923706055, "chaser_partner_vision": 0.9457500576972961, "chaser_return": 23.222501754760742, "chaser_subspace_norm": 34.96734619140625, "collisions_per_episode": 21.975000381469727, "entropy": 0.192807175219059, "explorer_grad_norm": 1.1616730093955994, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.9457500576972961, "explorer_return": 23.762500762939453, "explorer_subspace_norm": 49.27303695678711, "final_distance": 1.7601003646850586, "max_param_abs": 0.803512454032898, "policy_loss": -0.004968154602683406, "value_loss": 5.504559516906738}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:36.946485+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981216.946499, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.008178996266959, "elapsed_seconds": 3629.444386561052, "env_steps_per_second": 10469.92210176983, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1143031011371651, "eta_seconds": 4011.4911640937944, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0008471140095664964, "chaser_grad_norm": 0.24173415452241898, "chaser_new_fields": 47.35000228881836, "chaser_partner_vision": 0.9597500562667847, "chaser_return": 20.920000076293945, "chaser_subspace_norm": 34.23305130004883, "collisions_per_episode": 20.200000762939453, "entropy": 0.21809017285704613, "explorer_grad_norm": 0.2703899219632149, "explorer_new_fields": 58.10000228881836, "explorer_partner_vision": 0.9597500562667847, "explorer_return": 19.512500762939453, "explorer_subspace_norm": 44.29401397705078, "final_distance": 1.7585417032241821, "max_param_abs": 0.8384868502616882, "policy_loss": -0.006113522681410188, "value_loss": 5.599805235862732}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:47.004057+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981407.0040686, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.062454140260557, "elapsed_seconds": 3824.834904938005, "env_steps_per_second": 10457.967727798788, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.062454140260557, "eta_seconds": 3824.834904938005, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0007500614433881836, "chaser_grad_norm": 0.6011999174952507, "chaser_new_fields": 48.17499923706055, "chaser_partner_vision": 0.9185000658035278, "chaser_return": 21.000001907348633, "chaser_subspace_norm": 33.46017837524414, "collisions_per_episode": 20.05000114440918, "entropy": 0.16332949325442314, "explorer_grad_norm": 0.5066968724131584, "explorer_new_fields": 65.6500015258789, "explorer_partner_vision": 0.9185000658035278, "explorer_return": 31.662500381469727, "explorer_subspace_norm": 55.58749008178711, "final_distance": 2.09584379196167, "max_param_abs": 0.8596733212471008, "policy_loss": -0.004008793593847315, "value_loss": 5.301097393035889}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:40:02.394578+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981602.3945928, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1158693265263961, "elapsed_seconds": 4017.129575495026, "env_steps_per_second": 10455.226601652348, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.009596057333406, "eta_seconds": 3634.5458064002614, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006658367979857083, "chaser_grad_norm": 0.6580353677272797, "chaser_new_fields": 50.150001525878906, "chaser_partner_vision": 0.9515000581741333, "chaser_return": 18.19499969482422, "chaser_subspace_norm": 35.56165313720703, "collisions_per_episode": 17.05000114440918, "entropy": 0.14254750311374664, "explorer_grad_norm": 0.5595763102173805, "explorer_new_fields": 64.2249984741211, "explorer_partner_vision": 0.9515000581741333, "explorer_return": 31.700000762939453, "explorer_subspace_norm": 61.351722717285156, "final_distance": 1.9936646223068237, "max_param_abs": 0.8741118311882019, "policy_loss": -0.0032291742903147735, "value_loss": 5.351441740989685}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:14.689248+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981794.689261, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1693130982241644, "elapsed_seconds": 4209.527153606992, "env_steps_per_second": 10452.480384239352, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9567107167288619, "eta_seconds": 3444.158580223903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00033266255371600106, "chaser_grad_norm": 0.44620686024427414, "chaser_new_fields": 52.650001525878906, "chaser_partner_vision": 0.9292500615119934, "chaser_return": 20.850000381469727, "chaser_subspace_norm": 36.24545669555664, "collisions_per_episode": 19.350000381469727, "entropy": 0.16571272164583206, "explorer_grad_norm": 0.2641577273607254, "explorer_new_fields": 62.22500228881836, "explorer_partner_vision": 0.9292500615119934, "explorer_return": 25.48750114440918, "explorer_subspace_norm": 56.305633544921875, "final_distance": 1.6105679273605347, "max_param_abs": 0.8802231550216675, "policy_loss": -0.0034084307788901924, "value_loss": 5.08160924911499}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:27.086826+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981987.0868385, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2213211118661114, "elapsed_seconds": 4396.756002718001, "env_steps_per_second": 10462.258986298892, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.902715604422778, "eta_seconds": 3249.7761759220007, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00037514353079659557, "chaser_grad_norm": 0.27753085270524025, "chaser_new_fields": 49.42499923706055, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 26.315000534057617, "chaser_subspace_norm": 35.146785736083984, "collisions_per_episode": 24.950000762939453, "entropy": 0.16095273569226265, "explorer_grad_norm": 1.1096102595329285, "explorer_new_fields": 61.0, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 20.962499618530273, "explorer_subspace_norm": 58.78330993652344, "final_distance": 1.447013258934021, "max_param_abs": 0.9043868780136108, "policy_loss": -0.003552170393056997, "value_loss": 5.243370056152344}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:34.315673+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982174.3156843, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2738835591983257, "elapsed_seconds": 4585.9808131139725, "env_steps_per_second": 10466.681383127516, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.849255706132217, "eta_seconds": 3057.3205420759814, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0004910899558758786, "chaser_grad_norm": 0.24976231902837753, "chaser_new_fields": 51.275001525878906, "chaser_partner_vision": 0.9427500367164612, "chaser_return": 21.220001220703125, "chaser_subspace_norm": 34.35043716430664, "collisions_per_episode": 20.05000114440918, "entropy": 0.1522003933787346, "explorer_grad_norm": 0.6593773886561394, "explorer_new_fields": 65.8499984741211, "explorer_partner_vision": 0.9427500367164612, "explorer_return": 29.375, "explorer_subspace_norm": 57.87527847290039, "final_distance": 1.6798559427261353, "max_param_abs": 0.908561110496521, "policy_loss": -0.002972714798059606, "value_loss": 4.688760042190552}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:43.540485+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982363.5404994, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3257802743508365, "elapsed_seconds": 4772.808987663011, "env_steps_per_second": 10476.011114050956, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7954681646105018, "eta_seconds": 2863.6853925978066, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005639817899458777, "chaser_grad_norm": 0.4619976505637169, "chaser_new_fields": 51.900001525878906, "chaser_partner_vision": 0.9037500619888306, "chaser_return": 18.92500114440918, "chaser_subspace_norm": 33.58144760131836, "collisions_per_episode": 17.5, "entropy": 0.2394341193139553, "explorer_grad_norm": 0.44818028807640076, "explorer_new_fields": 60.275001525878906, "explorer_partner_vision": 0.9037500619888306, "explorer_return": 25.73750114440918, "explorer_subspace_norm": 66.47223663330078, "final_distance": 1.6649030447006226, "max_param_abs": 0.9298518896102905, "policy_loss": -0.005077735383995474, "value_loss": 5.802935600280762}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:50.368658+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982550.368668, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3778280866766646, "elapsed_seconds": 4960.181112035993, "env_steps_per_second": 10483.488168168056, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7419074312874349, "eta_seconds": 2670.8667526347654, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0003779172970928002, "chaser_grad_norm": 0.3904053717851639, "chaser_new_fields": 49.125, "chaser_partner_vision": 0.9535000324249268, "chaser_return": 21.14750099182129, "chaser_subspace_norm": 35.73765182495117, "collisions_per_episode": 20.225000381469727, "entropy": 0.21817343682050705, "explorer_grad_norm": 0.46166517585515976, "explorer_new_fields": 62.72500228881836, "explorer_partner_vision": 0.9535000324249268, "explorer_return": 25.200000762939453, "explorer_subspace_norm": 64.24727630615234, "final_distance": 1.5540331602096558, "max_param_abs": 0.9525908827781677, "policy_loss": -0.005648291538477679, "value_loss": 5.374347925186157}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:58:57.740784+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982737.7407978, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.429761035669718, "elapsed_seconds": 5147.139728410984, "env_steps_per_second": 10491.2636627937, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6884034616187531, "eta_seconds": 2478.252461827511, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0006776578440517156, "chaser_grad_norm": 0.7877217270433903, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9230000376701355, "chaser_return": 19.5625, "chaser_subspace_norm": 34.94178009033203, "collisions_per_episode": 18.075000762939453, "entropy": 0.21015918254852295, "explorer_grad_norm": 0.5025315135717392, "explorer_new_fields": 61.20000076293945, "explorer_partner_vision": 0.9230000376701355, "explorer_return": 26.200000762939453, "explorer_subspace_norm": 76.7821273803711, "final_distance": 1.7141870260238647, "max_param_abs": 0.9319895505905151, "policy_loss": -0.005408528198024953, "value_loss": 5.4712700843811035}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:04.699399+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982924.6994116, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.481721305380569, "elapsed_seconds": 5334.196699370048, "env_steps_per_second": 10498.300523228441, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6350234165916724, "eta_seconds": 2286.0842997300206, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0005930449717629349, "chaser_grad_norm": 0.3295997902750969, "chaser_new_fields": 48.85000228881836, "chaser_partner_vision": 0.9547500610351562, "chaser_return": 20.40500259399414, "chaser_subspace_norm": 39.16714859008789, "collisions_per_episode": 19.350000381469727, "entropy": 0.17486269399523735, "explorer_grad_norm": 0.2155264914035797, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.9547500610351562, "explorer_return": 23.162500381469727, "explorer_subspace_norm": 73.7138671875, "final_distance": 1.6524208784103394, "max_param_abs": 0.9965690970420837, "policy_loss": -0.0041415339447574695, "value_loss": 5.5186097621917725}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:11.756371+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983111.7563858, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5357093489113807, "elapsed_seconds": 5528.5536560809705, "env_steps_per_second": 10490.989797341408, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5825104426905238, "eta_seconds": 2097.0375936858854, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00145025184740466, "chaser_grad_norm": 0.6275696009397507, "chaser_new_fields": 46.400001525878906, "chaser_partner_vision": 0.9120000600814819, "chaser_return": 21.117502212524414, "chaser_subspace_norm": 37.609779357910156, "collisions_per_episode": 20.725000381469727, "entropy": 0.2002970427274704, "explorer_grad_norm": 0.2548380345106125, "explorer_new_fields": 56.97500228881836, "explorer_partner_vision": 0.9120000600814819, "explorer_return": 18.200000762939453, "explorer_subspace_norm": 69.76809692382812, "final_distance": 1.9717777967453003, "max_param_abs": 0.9967555999755859, "policy_loss": -0.006285025612451811, "value_loss": 5.378010988235474}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:26.113329+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983306.1133425, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5892785856830436, "elapsed_seconds": 5721.4029084589565, "env_steps_per_second": 10486.938424016851, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5297595285610145, "eta_seconds": 1907.1343028196522, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0012948946490425328, "chaser_grad_norm": 0.4338780380785465, "chaser_new_fields": 51.775001525878906, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 18.270002365112305, "chaser_subspace_norm": 39.203041076660156, "collisions_per_episode": 17.100000381469727, "entropy": 0.1832578182220459, "explorer_grad_norm": 0.4555763155221939, "explorer_new_fields": 60.17499923706055, "explorer_partner_vision": 0.9455000162124634, "explorer_return": 24.587499618530273, "explorer_subspace_norm": 88.89515686035156, "final_distance": 1.6995115280151367, "max_param_abs": 0.9677121639251709, "policy_loss": -0.0040612543104217025, "value_loss": 4.840352773666382}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:38.962580+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983498.962594, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6421395163169492, "elapsed_seconds": 5911.702258741017, "env_steps_per_second": 10487.67297242128, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47675018215653364, "eta_seconds": 1716.300655763521, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00013187378772006042, "chaser_grad_norm": 0.21623343229293823, "chaser_new_fields": 50.70000076293945, "chaser_partner_vision": 0.9292500615119934, "chaser_return": 18.787500381469727, "chaser_subspace_norm": 41.27761459350586, "collisions_per_episode": 17.725000381469727, "entropy": 0.17795084789395332, "explorer_grad_norm": 0.800887368619442, "explorer_new_fields": 59.75, "explorer_partner_vision": 0.9292500615119934, "explorer_return": 23.9375, "explorer_subspace_norm": 87.47830963134766, "final_distance": 1.6985321044921875, "max_param_abs": 0.9805355668067932, "policy_loss": -0.004621412475593445, "value_loss": 4.998327136039734}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:14:49.261931+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983689.2619443, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6949500843613836, "elapsed_seconds": 6101.820303700981, "env_steps_per_second": 10488.673349030227, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4237375210903459, "eta_seconds": 1525.4550759252452, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00011181542643168285, "chaser_grad_norm": 0.5658168196678162, "chaser_new_fields": 51.07500076293945, "chaser_partner_vision": 0.9100000262260437, "chaser_return": 19.767499923706055, "chaser_subspace_norm": 47.530330657958984, "collisions_per_episode": 18.575000762939453, "entropy": 0.13123946636915207, "explorer_grad_norm": 0.7169899940490723, "explorer_new_fields": 58.875, "explorer_partner_vision": 0.9100000262260437, "explorer_return": 23.774999618530273, "explorer_subspace_norm": 97.75777435302734, "final_distance": 1.9700640439987183, "max_param_abs": 1.0421918630599976, "policy_loss": -0.0032755948226039777, "value_loss": 5.5283859968185425}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:17:59.379975+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983879.37999, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7486499368844346, "elapsed_seconds": 6295.139772783965, "env_steps_per_second": 10484.278726477289, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3709257441876074, "eta_seconds": 1335.3326790753865, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005274816343927569, "chaser_grad_norm": 0.8685781061649323, "chaser_new_fields": 47.775001525878906, "chaser_partner_vision": 0.9335000514984131, "chaser_return": 24.697500228881836, "chaser_subspace_norm": 41.509193420410156, "collisions_per_episode": 23.67500114440918, "entropy": 0.2099326066672802, "explorer_grad_norm": 0.5854717269539833, "explorer_new_fields": 56.150001525878906, "explorer_partner_vision": 0.9335000514984131, "explorer_return": 14.512499809265137, "explorer_subspace_norm": 89.04560089111328, "final_distance": 1.5808110237121582, "max_param_abs": 1.1267714500427246, "policy_loss": -0.006268294434065069, "value_loss": 5.931675434112549}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:12.699446+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984072.6994596, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8026355679736137, "elapsed_seconds": 6489.488044705009, "env_steps_per_second": 10478.48451704653, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31811215905416707, "eta_seconds": 1145.2037725950015, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0006393396286394903, "chaser_grad_norm": 0.4078418090939522, "chaser_new_fields": 52.875, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 21.98750114440918, "chaser_subspace_norm": 42.20634841918945, "collisions_per_episode": 20.42500114440918, "entropy": 0.1676313653588295, "explorer_grad_norm": 0.9484191983938217, "explorer_new_fields": 60.07500076293945, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 21.3125, "explorer_subspace_norm": 109.76657104492188, "final_distance": 1.4780539274215698, "max_param_abs": 1.1382524967193604, "policy_loss": -0.004351576237473509, "value_loss": 5.102489709854126}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:27.047718+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984267.0477335, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.856240285412512, "elapsed_seconds": 6682.465027485043, "env_steps_per_second": 10475.176407521674, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26517718363035886, "eta_seconds": 954.6378610692918, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00026934010489432225, "chaser_grad_norm": 1.392246589064598, "chaser_new_fields": 51.75, "chaser_partner_vision": 0.9632500410079956, "chaser_return": 19.787500381469727, "chaser_subspace_norm": 46.96171188354492, "collisions_per_episode": 18.42500114440918, "entropy": 0.11750281229615211, "explorer_grad_norm": 0.36127040535211563, "explorer_new_fields": 57.04999923706055, "explorer_partner_vision": 0.9632500410079956, "explorer_return": 20.17500114440918, "explorer_subspace_norm": 91.96868896484375, "final_distance": 1.5381594896316528, "max_param_abs": 1.1600592136383057, "policy_loss": -0.001832261232191712, "value_loss": 5.648775577545166}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:40.024698+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984460.02471, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9098502091799876, "elapsed_seconds": 6875.460753047955, "env_steps_per_second": 10472.02545197887, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21220557879777638, "eta_seconds": 763.940083671995, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": -0.00026527807438861406, "chaser_grad_norm": 0.5024544820189476, "chaser_new_fields": 50.60000228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.29250144958496, "chaser_subspace_norm": 43.94972229003906, "collisions_per_episode": 18.975000381469727, "entropy": 0.15825418755412102, "explorer_grad_norm": 1.1086283177137375, "explorer_new_fields": 56.95000076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 19.712499618530273, "explorer_subspace_norm": 88.17216491699219, "final_distance": 1.569293737411499, "max_param_abs": 1.240469217300415, "policy_loss": -0.003059185811784104, "value_loss": 4.676573038101196}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:30:53.020426+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984653.0204408, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9630766453133481, "elapsed_seconds": 7067.075923128054, "env_steps_per_second": 10471.091694066004, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1591683766470282, "eta_seconds": 573.0061559293016, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": -0.00013887470340190333, "chaser_grad_norm": 0.6151726841926575, "chaser_new_fields": 50.75, "chaser_partner_vision": 0.9702500700950623, "chaser_return": 20.530000686645508, "chaser_subspace_norm": 45.46480178833008, "collisions_per_episode": 19.600000381469727, "entropy": 0.1452293060719967, "explorer_grad_norm": 0.7099784910678864, "explorer_new_fields": 60.70000076293945, "explorer_partner_vision": 0.9702500700950623, "explorer_return": 22.325000762939453, "explorer_subspace_norm": 80.15692901611328, "final_distance": 1.5210155248641968, "max_param_abs": 1.2077066898345947, "policy_loss": -0.0029535421817570295, "value_loss": 5.174638628959656}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:04.635594+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984844.6356046, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0153305566952784, "elapsed_seconds": 7255.190004103002, "env_steps_per_second": 10475.259773626878, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1060700292997515, "eta_seconds": 381.8521054791054, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0003448885976524707, "chaser_grad_norm": 0.35716934129595757, "chaser_new_fields": 51.95000076293945, "chaser_partner_vision": 0.9380000233650208, "chaser_return": 21.41499900817871, "chaser_subspace_norm": 46.69165802001953, "collisions_per_episode": 19.850000381469727, "entropy": 0.1385413333773613, "explorer_grad_norm": 0.4824305921792984, "explorer_new_fields": 58.07500076293945, "explorer_partner_vision": 0.9380000233650208, "explorer_return": 21.225000381469727, "explorer_subspace_norm": 84.86427307128906, "final_distance": 1.7785929441452026, "max_param_abs": 1.1600998640060425, "policy_loss": -0.0024347068886285683, "value_loss": 4.832224130630493}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:12.749676+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985032.7496898, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.069018816757501, "elapsed_seconds": 7448.467740327003, "env_steps_per_second": 10471.952449722987, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.053051764532243605, "eta_seconds": 190.986352316077, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00015810847441066755, "chaser_grad_norm": 0.22432011365890503, "chaser_new_fields": 49.97500228881836, "chaser_partner_vision": 0.9425000548362732, "chaser_return": 18.80000114440918, "chaser_subspace_norm": 44.65083694458008, "collisions_per_episode": 17.950000762939453, "entropy": 0.139775563031435, "explorer_grad_norm": 0.6975048556923866, "explorer_new_fields": 61.04999923706055, "explorer_partner_vision": 0.9425000548362732, "explorer_return": 24.950000762939453, "explorer_subspace_norm": 106.41658020019531, "final_distance": 1.5926132202148438, "max_param_abs": 1.1949032545089722, "policy_loss": -0.004057333317586398, "value_loss": 5.201714992523193}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:26.027411+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985226.027422, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1203027296616668, "elapsed_seconds": 7633.089826782001, "env_steps_per_second": 10480.683683206022, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:30.649499+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985410.649513, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1203057210158374, "elapsed_seconds": 7633.100595657015, "env_steps_per_second": 10480.668896924715, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.002805363851033249, "chaser_grad_norm": 0.188689436763525, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 23.260000228881836, "chaser_subspace_norm": 46.499481201171875, "collisions_per_episode": 21.450000762939453, "entropy": 0.1341038942337036, "explorer_grad_norm": 1.1695344895124435, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 105.87859344482422, "final_distance": 1.655848741531372, "max_param_abs": 1.2022045850753784, "policy_loss": -0.003802475623324142, "value_loss": 4.788339614868164}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:30.660267+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985410.6602778, "update": 20000} diff --git a/social/seed_0000/attempt_01/status.json b/social/seed_0000/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..6aef1aa2adf5376ccd4b89b98fdc403303640e83 --- /dev/null +++ b/social/seed_0000/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1203057210158374, + "elapsed_seconds": 7633.100595657015, + "env_steps_per_second": 10480.668896924715, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.002805363851033249, + "chaser_grad_norm": 0.188689436763525, + "chaser_new_fields": 52.92499923706055, + "chaser_partner_vision": 0.9490000605583191, + "chaser_return": 23.260000228881836, + "chaser_subspace_norm": 46.499481201171875, + "collisions_per_episode": 21.450000762939453, + "entropy": 0.1341038942337036, + "explorer_grad_norm": 1.1695344895124435, + "explorer_new_fields": 65.55000305175781, + "explorer_partner_vision": 0.9490000605583191, + "explorer_return": 31.037500381469727, + "explorer_subspace_norm": 105.87859344482422, + "final_distance": 1.655848741531372, + "max_param_abs": 1.2022045850753784, + "policy_loss": -0.003802475623324142, + "value_loss": 4.788339614868164 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:30.660267+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985410.6602778, + "update": 20000 +} diff --git a/social/seed_0001/attempt_01/checkpoints/latest.safetensors b/social/seed_0001/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a8895c4849a5b2652a559f8734d98256393d9709 --- /dev/null +++ b/social/seed_0001/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:099cb185f62dd1f3f5f4d2842a1b0d283ade548522932f999c600e73c1ff44a8 +size 11460400 diff --git a/social/seed_0001/attempt_01/config.json b/social/seed_0001/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..97e3600fa5ecbd2bbc8e3fd75592f92e5b5858da --- /dev/null +++ b/social/seed_0001/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0001.attempt_01", + "seed": 1, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0001/attempt_01/metrics.jsonl b/social/seed_0001/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..c5aff557a8323e492dc2e704832c552792721805 --- /dev/null +++ b/social/seed_0001/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005559027094730279, "elapsed_seconds": 2.0012497541029006, "env_steps_per_second": 1998.7510263520699, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 11.117498286751086, "eta_seconds": 40022.99383230391, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -2.6168578176624657e-05, "chaser_grad_norm": 0.21696332097053528, "chaser_new_fields": 34.54999923706055, "chaser_partner_vision": 0.31975001096725464, "chaser_return": -1.942500352859497, "chaser_subspace_norm": 0.0024611102417111397, "collisions_per_episode": 1.0750000476837158, "entropy": 2.7724967002868652, "explorer_grad_norm": 0.12145722843706608, "explorer_new_fields": 36.20000076293945, "explorer_partner_vision": 0.31975001096725464, "explorer_return": 3.5, "explorer_subspace_norm": 0.0023492525797337294, "final_distance": 5.040530681610107, "max_param_abs": 0.15427561104297638, "policy_loss": -0.00072323851910161, "value_loss": 2.967313230037689}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.758885+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.7589002, "update": 1} +{"device": "cuda", "elapsed_hours": 0.053880469897509706, "elapsed_seconds": 193.96969163103495, "env_steps_per_second": 10310.8892073941, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1013383260028786, "eta_seconds": 7564.817973610363, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0010252506795414007, "chaser_grad_norm": 0.333268154412508, "chaser_new_fields": 51.04999923706055, "chaser_partner_vision": 0.8815000653266907, "chaser_return": 11.6225004196167, "chaser_subspace_norm": 17.339336395263672, "collisions_per_episode": 10.824999809265137, "entropy": 0.549514040350914, "explorer_grad_norm": 0.30565332621335983, "explorer_new_fields": 61.97500228881836, "explorer_partner_vision": 0.8815000653266907, "explorer_return": 33.875, "explorer_subspace_norm": 11.0411958694458, "final_distance": 2.112823963165283, "max_param_abs": 0.31764182448387146, "policy_loss": -0.006427154921922806, "value_loss": 5.539729833602905}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:31.727326+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977971.7273388, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10743285710891036, "elapsed_seconds": 386.7582855920773, "env_steps_per_second": 10342.377006549486, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0412242850692968, "eta_seconds": 7348.407426249469, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.000996291420352069, "chaser_grad_norm": 0.3361887261271477, "chaser_new_fields": 51.125, "chaser_partner_vision": 0.8857500553131104, "chaser_return": 13.367500305175781, "chaser_subspace_norm": 24.21672248840332, "collisions_per_episode": 12.324999809265137, "entropy": 0.3819955810904503, "explorer_grad_norm": 0.37394415587186813, "explorer_new_fields": 59.85000228881836, "explorer_partner_vision": 0.8857500553131104, "explorer_return": 29.600000381469727, "explorer_subspace_norm": 16.2431640625, "final_distance": 1.9537521600723267, "max_param_abs": 0.3494575619697571, "policy_loss": -0.006638709723958414, "value_loss": 5.393805027008057}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:44.515919+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978164.5159342, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16057416942834, "elapsed_seconds": 578.067009942024, "env_steps_per_second": 10379.419508132383, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9804147562828598, "eta_seconds": 7129.493122618295, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00014566476395883043, "chaser_grad_norm": 0.43530627712607384, "chaser_new_fields": 48.75, "chaser_partner_vision": 0.8400000333786011, "chaser_return": 20.125, "chaser_subspace_norm": 30.273305892944336, "collisions_per_episode": 18.950000762939453, "entropy": 0.2900804206728935, "explorer_grad_norm": 0.7670780718326569, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.8400000333786011, "explorer_return": 25.5, "explorer_subspace_norm": 23.51873779296875, "final_distance": 2.349271535873413, "max_param_abs": 0.37496188282966614, "policy_loss": -0.0017435111670058845, "value_loss": 5.994792580604553}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:55.824642+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978355.8246543, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21239668673473514, "elapsed_seconds": 764.6280722450465, "env_steps_per_second": 10462.603048971207, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9115701806126162, "eta_seconds": 6881.6526502054185, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0006213136221422388, "chaser_grad_norm": 0.43084486573934555, "chaser_new_fields": 51.525001525878906, "chaser_partner_vision": 0.8582500219345093, "chaser_return": 13.99000072479248, "chaser_subspace_norm": 29.42274284362793, "collisions_per_episode": 13.0, "entropy": 0.2611316740512848, "explorer_grad_norm": 0.6707221195101738, "explorer_new_fields": 64.375, "explorer_partner_vision": 0.8582500219345093, "explorer_return": 35.07500076293945, "explorer_subspace_norm": 26.8638916015625, "final_distance": 2.146491289138794, "max_param_abs": 0.4319896996021271, "policy_loss": -0.004898147379513773, "value_loss": 5.452469825744629}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:02.385706+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978542.3857198, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26676166866889817, "elapsed_seconds": 960.3420072080335, "env_steps_per_second": 10412.956972560876, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8673316806822873, "eta_seconds": 6722.394050456234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0006523600313632083, "chaser_grad_norm": 0.4621363505721092, "chaser_new_fields": 52.25, "chaser_partner_vision": 0.8417500257492065, "chaser_return": 15.852500915527344, "chaser_subspace_norm": 33.860740661621094, "collisions_per_episode": 14.625, "entropy": 0.2149227149784565, "explorer_grad_norm": 0.435776986181736, "explorer_new_fields": 67.625, "explorer_partner_vision": 0.8417500257492065, "explorer_return": 38.38750076293945, "explorer_subspace_norm": 29.73369598388672, "final_distance": 2.442427635192871, "max_param_abs": 0.44966426491737366, "policy_loss": -0.0037778343553682703, "value_loss": 5.461779832839966}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:18.099639+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978738.0996518, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3206045408947264, "elapsed_seconds": 1154.176347221015, "env_steps_per_second": 10397.024708479927, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8167590650701164, "eta_seconds": 6540.332634252419, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0003632399116328955, "chaser_grad_norm": 0.4582560732960701, "chaser_new_fields": 54.95000076293945, "chaser_partner_vision": 0.874750018119812, "chaser_return": 16.84000015258789, "chaser_subspace_norm": 37.34330749511719, "collisions_per_episode": 15.0, "entropy": 0.1612669862806797, "explorer_grad_norm": 0.6384056285023689, "explorer_new_fields": 69.4000015258789, "explorer_partner_vision": 0.874750018119812, "explorer_return": 40.11249923706055, "explorer_subspace_norm": 33.11635208129883, "final_distance": 1.6558111906051636, "max_param_abs": 0.4803593158721924, "policy_loss": -0.0036103047044022674, "value_loss": 5.186581254005432}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:31.933980+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978931.9339945, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.373498961604459, "elapsed_seconds": 1344.5962617760524, "env_steps_per_second": 10412.04739146579, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7607808189924496, "eta_seconds": 6338.8109483728185, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.001289308286482438, "chaser_grad_norm": 0.5507540702819824, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.9237500429153442, "chaser_return": 18.347501754760742, "chaser_subspace_norm": 35.409080505371094, "collisions_per_episode": 17.125, "entropy": 0.20060503110289574, "explorer_grad_norm": 0.4468846917152405, "explorer_new_fields": 62.97500228881836, "explorer_partner_vision": 0.9237500429153442, "explorer_return": 30.012500762939453, "explorer_subspace_norm": 33.27703857421875, "final_distance": 1.9853023290634155, "max_param_abs": 0.48706626892089844, "policy_loss": -0.00562318036146614, "value_loss": 5.48871648311615}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:42.353894+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979122.3539052, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4265518300014082, "elapsed_seconds": 1535.5865880050696, "env_steps_per_second": 10419.471050985227, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.706207320005633, "eta_seconds": 6142.346352020279, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0005951074044237004, "chaser_grad_norm": 0.5060320645570755, "chaser_new_fields": 51.45000076293945, "chaser_partner_vision": 0.8985000252723694, "chaser_return": 14.882501602172852, "chaser_subspace_norm": 34.31396484375, "collisions_per_episode": 13.875, "entropy": 0.2120460756123066, "explorer_grad_norm": 0.45205342024564743, "explorer_new_fields": 61.625, "explorer_partner_vision": 0.8985000252723694, "explorer_return": 30.36250114440918, "explorer_subspace_norm": 35.848323822021484, "final_distance": 1.9331448078155518, "max_param_abs": 0.5321258306503296, "policy_loss": -0.005638409091162089, "value_loss": 5.384594798088074}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:53.344223+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979313.3442373, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4806323333555742, "elapsed_seconds": 1730.276400080067, "env_steps_per_second": 10402.96220833103, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6555113704469777, "eta_seconds": 5959.84093360912, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0007196299260487926, "chaser_grad_norm": 0.4056941270828247, "chaser_new_fields": 48.625, "chaser_partner_vision": 0.8997500538825989, "chaser_return": 17.020002365112305, "chaser_subspace_norm": 33.130062103271484, "collisions_per_episode": 16.399999618530273, "entropy": 0.19867215678095818, "explorer_grad_norm": 0.3142911419272423, "explorer_new_fields": 61.10000228881836, "explorer_partner_vision": 0.8997500538825989, "explorer_return": 27.637500762939453, "explorer_subspace_norm": 40.459556579589844, "final_distance": 1.7062615156173706, "max_param_abs": 0.5415729284286499, "policy_loss": -0.005311255770273249, "value_loss": 5.52983820438385}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:05:08.034034+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979508.0340478, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5341139293369552, "elapsed_seconds": 1922.810145613039, "env_steps_per_second": 10401.442932694486, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6023417880108657, "eta_seconds": 5768.430436839117, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00046516035691990254, "chaser_grad_norm": 1.3175133019685745, "chaser_new_fields": 49.42499923706055, "chaser_partner_vision": 0.8905000686645508, "chaser_return": 17.9325008392334, "chaser_subspace_norm": 32.691673278808594, "collisions_per_episode": 17.125, "entropy": 0.23665303364396095, "explorer_grad_norm": 0.5993022322654724, "explorer_new_fields": 62.875, "explorer_partner_vision": 0.8905000686645508, "explorer_return": 29.30000114440918, "explorer_subspace_norm": 38.85688781738281, "final_distance": 1.773389458656311, "max_param_abs": 0.5497462749481201, "policy_loss": -0.005516570260752474, "value_loss": 5.747911334037781}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:20.567779+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979700.5677931, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5872149470794507, "elapsed_seconds": 2113.973809486022, "env_steps_per_second": 10406.940663729858, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5481121332094607, "eta_seconds": 5573.203679554059, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0013853114010555778, "chaser_grad_norm": 0.26142775267362595, "chaser_new_fields": 49.375, "chaser_partner_vision": 0.9162500500679016, "chaser_return": 18.277502059936523, "chaser_subspace_norm": 33.60045623779297, "collisions_per_episode": 17.274999618530273, "entropy": 0.20493536815047264, "explorer_grad_norm": 0.21770518273115158, "explorer_new_fields": 62.275001525878906, "explorer_partner_vision": 0.9162500500679016, "explorer_return": 28.962499618530273, "explorer_subspace_norm": 44.72954177856445, "final_distance": 2.045193910598755, "max_param_abs": 0.5616775155067444, "policy_loss": -0.006837340514174173, "value_loss": 5.578496217727661}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:31.731442+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979891.7314546, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6407613792408503, "elapsed_seconds": 2306.7409652670613, "env_steps_per_second": 10404.289151392175, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4951098848953175, "eta_seconds": 5382.395585623143, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00069671257208892, "chaser_grad_norm": 0.5549150705337524, "chaser_new_fields": 50.20000076293945, "chaser_partner_vision": 0.9447500705718994, "chaser_return": 17.087499618530273, "chaser_subspace_norm": 36.431636810302734, "collisions_per_episode": 16.17500114440918, "entropy": 0.1903597116470337, "explorer_grad_norm": 0.4318744130432606, "explorer_new_fields": 61.400001525878906, "explorer_partner_vision": 0.9447500705718994, "explorer_return": 29.212499618530273, "explorer_subspace_norm": 50.51710891723633, "final_distance": 1.6914364099502563, "max_param_abs": 0.6074740290641785, "policy_loss": -0.004711625126656571, "value_loss": 5.509023189544678}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:44.498598+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980084.4986138, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6936579273891725, "elapsed_seconds": 2497.1685386010213, "env_steps_per_second": 10411.792235123175, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4406741568852044, "eta_seconds": 5186.426964786736, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003491099669616915, "chaser_grad_norm": 0.37194397300481796, "chaser_new_fields": 50.07500076293945, "chaser_partner_vision": 0.9342500567436218, "chaser_return": 19.880002975463867, "chaser_subspace_norm": 36.51939392089844, "collisions_per_episode": 18.649999618530273, "entropy": 0.1523255854845047, "explorer_grad_norm": 0.8375664204359055, "explorer_new_fields": 63.04999923706055, "explorer_partner_vision": 0.9342500567436218, "explorer_return": 28.6875, "explorer_subspace_norm": 45.561946868896484, "final_distance": 1.689591646194458, "max_param_abs": 0.6433732509613037, "policy_loss": -0.0035669016917232454, "value_loss": 5.049667119979858}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:54.926170+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980274.9261818, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7456941334319547, "elapsed_seconds": 2684.498880355037, "env_steps_per_second": 10430.252068608379, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3848605335164874, "eta_seconds": 4985.497920659354, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0009760428614704608, "chaser_grad_norm": 0.7625705525279045, "chaser_new_fields": 50.75, "chaser_partner_vision": 0.9472500681877136, "chaser_return": 19.232501983642578, "chaser_subspace_norm": 40.61727523803711, "collisions_per_episode": 18.125, "entropy": 0.14908970519900322, "explorer_grad_norm": 0.29900267720222473, "explorer_new_fields": 59.32500076293945, "explorer_partner_vision": 0.9472500681877136, "explorer_return": 24.825000762939453, "explorer_subspace_norm": 52.48163604736328, "final_distance": 1.4705226421356201, "max_param_abs": 0.7251474857330322, "policy_loss": -0.004909361024385106, "value_loss": 5.499907612800598}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:21:02.256514+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980462.256528, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.8006891609405608, "elapsed_seconds": 2882.4809793860186, "env_steps_per_second": 10407.700940455168, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3344819349009345, "eta_seconds": 4804.134965643364, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": -0.0004871895799434256, "chaser_grad_norm": 0.2261325791478157, "chaser_new_fields": 53.75, "chaser_partner_vision": 0.9417500495910645, "chaser_return": 16.350000381469727, "chaser_subspace_norm": 38.88923263549805, "collisions_per_episode": 14.949999809265137, "entropy": 0.13790288195014, "explorer_grad_norm": 1.448859840631485, "explorer_new_fields": 63.17499923706055, "explorer_partner_vision": 0.9417500495910645, "explorer_return": 31.17500114440918, "explorer_subspace_norm": 63.22129821777344, "final_distance": 1.808421015739441, "max_param_abs": 0.7265934348106384, "policy_loss": -0.0031175145157593676, "value_loss": 5.200205206871033}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:20.238612+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980660.2386246, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8543859176222274, "elapsed_seconds": 3075.7893034400186, "env_steps_per_second": 10403.833566951618, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2815788764333411, "eta_seconds": 4613.683955160028, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00019009302998007627, "chaser_grad_norm": 0.5216067247092724, "chaser_new_fields": 52.525001525878906, "chaser_partner_vision": 0.9530000686645508, "chaser_return": 22.537500381469727, "chaser_subspace_norm": 40.95054626464844, "collisions_per_episode": 21.075000762939453, "entropy": 0.11076595261693001, "explorer_grad_norm": 0.48997530341148376, "explorer_new_fields": 64.55000305175781, "explorer_partner_vision": 0.9530000686645508, "explorer_return": 27.325000762939453, "explorer_subspace_norm": 57.401737213134766, "final_distance": 1.6586898565292358, "max_param_abs": 0.7021496295928955, "policy_loss": -0.002399672576692602, "value_loss": 5.102726578712463}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:33.546937+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980853.546951, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9079500129658522, "elapsed_seconds": 3268.620046677068, "env_steps_per_second": 10401.94317922175, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.228402958718506, "eta_seconds": 4422.250651386621, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00078926079664976, "chaser_grad_norm": 0.534687414765358, "chaser_new_fields": 50.875, "chaser_partner_vision": 0.9417500495910645, "chaser_return": 24.649999618530273, "chaser_subspace_norm": 35.6802864074707, "collisions_per_episode": 23.149999618530273, "entropy": 0.18846818059682846, "explorer_grad_norm": 0.2650442235171795, "explorer_new_fields": 59.72500228881836, "explorer_partner_vision": 0.9417500495910645, "explorer_return": 20.100000381469727, "explorer_subspace_norm": 58.25392532348633, "final_distance": 1.7197293043136597, "max_param_abs": 0.7541922926902771, "policy_loss": -0.005141545460094932, "value_loss": 5.366925001144409}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:46.377679+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981046.3776908, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9599589397750161, "elapsed_seconds": 3455.852183190058, "env_steps_per_second": 10417.11221767848, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1732831486139086, "eta_seconds": 4223.819335010071, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00044918432325924407, "chaser_grad_norm": 0.4804118573665619, "chaser_new_fields": 43.025001525878906, "chaser_partner_vision": 0.9390000700950623, "chaser_return": 18.642501831054688, "chaser_subspace_norm": 32.35923385620117, "collisions_per_episode": 19.075000762939453, "entropy": 0.20543936640024185, "explorer_grad_norm": 0.6772683635354042, "explorer_new_fields": 57.10000228881836, "explorer_partner_vision": 0.9390000700950623, "explorer_return": 18.462499618530273, "explorer_subspace_norm": 63.37069320678711, "final_distance": 1.6161651611328125, "max_param_abs": 0.7707778811454773, "policy_loss": -0.00464805631106735, "value_loss": 5.449813604354858}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:53.609816+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981233.609831, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0129418162072479, "elapsed_seconds": 3646.590538346092, "env_steps_per_second": 10420.692863760587, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.119567270544853, "eta_seconds": 4030.4421739614704, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0004441034328336313, "chaser_grad_norm": 0.608376320451498, "chaser_new_fields": 45.85000228881836, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 24.822500228881836, "chaser_subspace_norm": 36.03782272338867, "collisions_per_episode": 24.17500114440918, "entropy": 0.1306472308933735, "explorer_grad_norm": 0.7583857923746109, "explorer_new_fields": 53.67499923706055, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 10.475000381469727, "explorer_subspace_norm": 70.65046691894531, "final_distance": 1.9893743991851807, "max_param_abs": 0.7582400441169739, "policy_loss": -0.0029265512963752016, "value_loss": 5.661183834075928}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:37:04.348174+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981424.3481905, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.066931471211396, "elapsed_seconds": 3840.9532963610254, "env_steps_per_second": 10414.08132660623, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.066931471211396, "eta_seconds": 3840.9532963610254, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0007254199601620659, "chaser_grad_norm": 0.5257556959986687, "chaser_new_fields": 49.35000228881836, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 19.89750099182129, "chaser_subspace_norm": 36.68754196166992, "collisions_per_episode": 19.075000762939453, "entropy": 0.18798119574785233, "explorer_grad_norm": 0.32703598588705063, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 20.075000762939453, "explorer_subspace_norm": 67.89421081542969, "final_distance": 1.6082544326782227, "max_param_abs": 0.7855800986289978, "policy_loss": -0.004314257098361196, "value_loss": 5.290697932243347}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:40:18.710930+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981618.710944, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1201572898491738, "elapsed_seconds": 4032.566243457026, "env_steps_per_second": 10415.203982859899, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0134756431968717, "eta_seconds": 3648.512315508738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.000932469552051618, "chaser_grad_norm": 0.42100925371050835, "chaser_new_fields": 52.625, "chaser_partner_vision": 0.9542500376701355, "chaser_return": 25.375, "chaser_subspace_norm": 36.46957015991211, "collisions_per_episode": 23.600000381469727, "entropy": 0.1380283273756504, "explorer_grad_norm": 0.3326633498072624, "explorer_new_fields": 64.80000305175781, "explorer_partner_vision": 0.9542500376701355, "explorer_return": 24.524999618530273, "explorer_subspace_norm": 73.58235168457031, "final_distance": 1.6721057891845703, "max_param_abs": 0.8458138108253479, "policy_loss": -0.004104187934194314, "value_loss": 4.450531721115112}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:30.323876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981810.3238866, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1733347408177925, "elapsed_seconds": 4224.005066944053, "env_steps_per_second": 10416.654171258544, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.960001151578194, "eta_seconds": 3456.0041456814984, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.000622664010686471, "chaser_grad_norm": 0.4324997365474701, "chaser_new_fields": 51.85000228881836, "chaser_partner_vision": 0.9410000443458557, "chaser_return": 22.150001525878906, "chaser_subspace_norm": 35.965126037597656, "collisions_per_episode": 20.75, "entropy": 0.1802142672240734, "explorer_grad_norm": 0.6823921948671341, "explorer_new_fields": 59.525001525878906, "explorer_partner_vision": 0.9410000443458557, "explorer_return": 21.67500114440918, "explorer_subspace_norm": 78.4558334350586, "final_distance": 1.6906040906906128, "max_param_abs": 0.8729985356330872, "policy_loss": -0.004779425607054577, "value_loss": 5.132414698600769}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:41.762700+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982001.7627144, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.225777449778624, "elapsed_seconds": 4412.798819203046, "env_steps_per_second": 10424.223238961895, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9060094194015916, "eta_seconds": 3261.6339098457297, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00044057945389197606, "chaser_grad_norm": 0.39360369369387627, "chaser_new_fields": 53.0, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 26.40250015258789, "chaser_subspace_norm": 38.70211410522461, "collisions_per_episode": 24.475000381469727, "entropy": 0.13198284804821014, "explorer_grad_norm": 0.602433331310749, "explorer_new_fields": 61.42499923706055, "explorer_partner_vision": 0.9455000162124634, "explorer_return": 19.287500381469727, "explorer_subspace_norm": 70.93849182128906, "final_distance": 1.700857400894165, "max_param_abs": 0.9139685034751892, "policy_loss": -0.0036093499865383905, "value_loss": 4.277888536453247}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:50.556452+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982190.5564632, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2781215595669668, "elapsed_seconds": 4601.237614441081, "env_steps_per_second": 10431.975920858986, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8520810397113113, "eta_seconds": 3067.491742960721, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00032651641323322433, "chaser_grad_norm": 0.5646921545267105, "chaser_new_fields": 45.22500228881836, "chaser_partner_vision": 0.940500020980835, "chaser_return": 26.09250259399414, "chaser_subspace_norm": 37.345882415771484, "collisions_per_episode": 25.125, "entropy": 0.17272819206118584, "explorer_grad_norm": 0.3577100932598114, "explorer_new_fields": 55.60000228881836, "explorer_partner_vision": 0.940500020980835, "explorer_return": 14.08750057220459, "explorer_subspace_norm": 76.94121551513672, "final_distance": 1.5656722784042358, "max_param_abs": 0.9353617429733276, "policy_loss": -0.005104706510901114, "value_loss": 5.135862112045288}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:58.995249+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982378.9952638, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.330080691172528, "elapsed_seconds": 4788.2904882211005, "env_steps_per_second": 10442.140075460526, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7980484147035167, "eta_seconds": 2872.9742929326603, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0017905708461372072, "chaser_grad_norm": 1.6469353064894676, "chaser_new_fields": 49.60000228881836, "chaser_partner_vision": 0.9575000405311584, "chaser_return": 26.42500114440918, "chaser_subspace_norm": 36.2133674621582, "collisions_per_episode": 25.0, "entropy": 0.17152965441346169, "explorer_grad_norm": 0.7303834557533264, "explorer_new_fields": 60.04999923706055, "explorer_partner_vision": 0.9575000405311584, "explorer_return": 18.61250114440918, "explorer_subspace_norm": 72.41524505615234, "final_distance": 1.6450576782226562, "max_param_abs": 0.9972871541976929, "policy_loss": -0.005229052831417391, "value_loss": 5.023063063621521}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:56:06.048120+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982566.0481315, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3829733306866707, "elapsed_seconds": 4978.703990472015, "env_steps_per_second": 10444.485171143915, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7446779472928228, "eta_seconds": 2680.840610254162, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0005653295019394378, "chaser_grad_norm": 0.7697694711387157, "chaser_new_fields": 46.125, "chaser_partner_vision": 0.9160000681877136, "chaser_return": 17.297500610351562, "chaser_subspace_norm": 39.29671859741211, "collisions_per_episode": 16.975000381469727, "entropy": 0.22161920741200447, "explorer_grad_norm": 0.6069164425134659, "explorer_new_fields": 54.70000076293945, "explorer_partner_vision": 0.9160000681877136, "explorer_return": 18.762500762939453, "explorer_subspace_norm": 84.57980346679688, "final_distance": 1.8015260696411133, "max_param_abs": 1.0585379600524902, "policy_loss": -0.0037898754263316903, "value_loss": 5.84531569480896}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:16.461626+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982756.4616416, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4364987528064133, "elapsed_seconds": 5171.395510103088, "env_steps_per_second": 10442.05570711097, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6916475476475323, "eta_seconds": 2489.9311715311164, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0013080395428446412, "chaser_grad_norm": 0.22935017943382263, "chaser_new_fields": 41.400001525878906, "chaser_partner_vision": 0.8992500305175781, "chaser_return": 23.57750129699707, "chaser_subspace_norm": 48.25534439086914, "collisions_per_episode": 23.67500114440918, "entropy": 0.2594703957438469, "explorer_grad_norm": 0.33775416016578674, "explorer_new_fields": 53.97500228881836, "explorer_partner_vision": 0.8992500305175781, "explorer_return": 14.25, "explorer_subspace_norm": 90.6865005493164, "final_distance": 2.188667058944702, "max_param_abs": 1.102486491203308, "policy_loss": -0.003948161364998981, "value_loss": 6.881443500518799}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:29.153144+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982949.153158, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4899892016083403, "elapsed_seconds": 5363.961125790025, "env_steps_per_second": 10440.045833060005, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6385668006892887, "eta_seconds": 2298.8404824814393, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0006209199249367359, "chaser_grad_norm": 0.2556704394519329, "chaser_new_fields": 46.60000228881836, "chaser_partner_vision": 0.9420000314712524, "chaser_return": 21.035001754760742, "chaser_subspace_norm": 50.046875, "collisions_per_episode": 20.450000762939453, "entropy": 0.1712327040731907, "explorer_grad_norm": 0.8527882248163223, "explorer_new_fields": 59.70000076293945, "explorer_partner_vision": 0.9420000314712524, "explorer_return": 21.5625, "explorer_subspace_norm": 96.96524810791016, "final_distance": 1.8160396814346313, "max_param_abs": 1.120617389678955, "policy_loss": -0.00446073596645391, "value_loss": 6.012434959411621}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:41.718759+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983141.718774, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5418140142355696, "elapsed_seconds": 5550.530451248051, "env_steps_per_second": 10449.451725277637, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5848260053996989, "eta_seconds": 2105.3736194389157, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0003183286846279998, "chaser_grad_norm": 0.6510180085897446, "chaser_new_fields": 49.22500228881836, "chaser_partner_vision": 0.9495000243186951, "chaser_return": 21.29250144958496, "chaser_subspace_norm": 48.05588912963867, "collisions_per_episode": 20.17500114440918, "entropy": 0.16617627814412117, "explorer_grad_norm": 0.600015789270401, "explorer_new_fields": 61.625, "explorer_partner_vision": 0.9495000243186951, "explorer_return": 25.67500114440918, "explorer_subspace_norm": 96.63275146484375, "final_distance": 1.6368759870529175, "max_param_abs": 1.1139692068099976, "policy_loss": -0.004752415399066123, "value_loss": 5.3756715059280396}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:48.288084+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983328.2880948, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.594882148113619, "elapsed_seconds": 5741.575733209029, "env_steps_per_second": 10450.092934063825, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5316273827045397, "eta_seconds": 1913.858577736343, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0015743686322318151, "chaser_grad_norm": 0.351873267441988, "chaser_new_fields": 50.525001525878906, "chaser_partner_vision": 0.9607500433921814, "chaser_return": 22.00750160217285, "chaser_subspace_norm": 51.991981506347656, "collisions_per_episode": 20.575000762939453, "entropy": 0.1728307232260704, "explorer_grad_norm": 0.43463999778032303, "explorer_new_fields": 57.22500228881836, "explorer_partner_vision": 0.9607500433921814, "explorer_return": 20.1875, "explorer_subspace_norm": 93.74605560302734, "final_distance": 1.7956188917160034, "max_param_abs": 1.138116478919983, "policy_loss": -0.005384587672537577, "value_loss": 5.726532697677612}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:59.333368+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983519.333384, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6485902026613863, "elapsed_seconds": 5934.924729580991, "env_steps_per_second": 10446.636280149965, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4786229620629831, "eta_seconds": 1723.0426634267392, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0013003932424249243, "chaser_grad_norm": 0.2659047693014145, "chaser_new_fields": 50.25, "chaser_partner_vision": 0.908750057220459, "chaser_return": 19.35249900817871, "chaser_subspace_norm": 48.25482177734375, "collisions_per_episode": 18.42500114440918, "entropy": 0.17284099385142326, "explorer_grad_norm": 0.28693175688385963, "explorer_new_fields": 58.900001525878906, "explorer_partner_vision": 0.908750057220459, "explorer_return": 21.412500381469727, "explorer_subspace_norm": 89.82017517089844, "final_distance": 1.7588703632354736, "max_param_abs": 1.1763365268707275, "policy_loss": -0.004858753234138646, "value_loss": 5.372735261917114}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:15:12.682363+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983712.682377, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.7025206530252923, "elapsed_seconds": 6129.074350891053, "env_steps_per_second": 10442.033549600454, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4256301632563231, "eta_seconds": 1532.2685877227632, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006692240439161168, "chaser_grad_norm": 0.22271914035081863, "chaser_new_fields": 50.275001525878906, "chaser_partner_vision": 0.9515000581741333, "chaser_return": 19.38250160217285, "chaser_subspace_norm": 49.52060317993164, "collisions_per_episode": 18.225000381469727, "entropy": 0.14475304633378983, "explorer_grad_norm": 0.5359522774815559, "explorer_new_fields": 56.95000076293945, "explorer_partner_vision": 0.9515000581741333, "explorer_return": 19.412500381469727, "explorer_subspace_norm": 103.75702667236328, "final_distance": 1.7306162118911743, "max_param_abs": 1.2046598196029663, "policy_loss": -0.003649979956463767, "value_loss": 5.052820086479187}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:26.831984+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983906.8319976, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7544221175097239, "elapsed_seconds": 6315.919623035006, "env_steps_per_second": 10449.784661490805, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.37215014613842634, "eta_seconds": 1339.7405260983348, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006272952446830793, "chaser_grad_norm": 0.7742678821086884, "chaser_new_fields": 49.42499923706055, "chaser_partner_vision": 0.9347500205039978, "chaser_return": 18.024999618530273, "chaser_subspace_norm": 50.26509475708008, "collisions_per_episode": 17.149999618530273, "entropy": 0.19025595486164093, "explorer_grad_norm": 0.7143794447183609, "explorer_new_fields": 57.20000076293945, "explorer_partner_vision": 0.9347500205039978, "explorer_return": 22.274999618530273, "explorer_subspace_norm": 84.62357330322266, "final_distance": 1.6148922443389893, "max_param_abs": 1.2372798919677734, "policy_loss": -0.004013667650054842, "value_loss": 5.5340269804000854}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:33.677255+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984093.6772668, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.808249243025849, "elapsed_seconds": 6509.697274893057, "env_steps_per_second": 10445.954263075486, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3191028075927969, "eta_seconds": 1148.7701073340688, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": -8.016964674228966e-05, "chaser_grad_norm": 0.40535828471183777, "chaser_new_fields": 48.625, "chaser_partner_vision": 0.9517500400543213, "chaser_return": 20.435001373291016, "chaser_subspace_norm": 58.668113708496094, "collisions_per_episode": 19.30000114440918, "entropy": 0.11981409229338169, "explorer_grad_norm": 0.7677905410528183, "explorer_new_fields": 61.150001525878906, "explorer_partner_vision": 0.9517500400543213, "explorer_return": 25.287500381469727, "explorer_subspace_norm": 92.68165588378906, "final_distance": 1.5604958534240723, "max_param_abs": 1.2742111682891846, "policy_loss": -0.003531392759644092, "value_loss": 4.820449590682983}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:47.454908+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984287.4549212, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.861714405151401, "elapsed_seconds": 6702.1718585450435, "env_steps_per_second": 10444.375566220726, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26595920073591445, "eta_seconds": 957.4531226492919, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.000846881658597809, "chaser_grad_norm": 0.3458448611199856, "chaser_new_fields": 55.25, "chaser_partner_vision": 0.9590000510215759, "chaser_return": 17.635000228881836, "chaser_subspace_norm": 51.43133544921875, "collisions_per_episode": 15.699999809265137, "entropy": 0.13018999621272087, "explorer_grad_norm": 0.44314559549093246, "explorer_new_fields": 60.125, "explorer_partner_vision": 0.9590000510215759, "explorer_return": 26.9375, "explorer_subspace_norm": 102.31692504882812, "final_distance": 1.5663975477218628, "max_param_abs": 1.271032452583313, "policy_loss": -0.002613898420706917, "value_loss": 5.384448409080505}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:59.929492+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984479.9295058, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9149907900491638, "elapsed_seconds": 6893.96684417699, "env_steps_per_second": 10443.914458453628, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21277675444990712, "eta_seconds": 765.9963160196656, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00020224497337117953, "chaser_grad_norm": 0.25990021601319313, "chaser_new_fields": 50.900001525878906, "chaser_partner_vision": 0.9177500605583191, "chaser_return": 20.420000076293945, "chaser_subspace_norm": 55.61945343017578, "collisions_per_episode": 19.100000381469727, "entropy": 0.12409074790775776, "explorer_grad_norm": 0.37726984545588493, "explorer_new_fields": 58.25, "explorer_partner_vision": 0.9177500605583191, "explorer_return": 20.287500381469727, "explorer_subspace_norm": 89.28801727294922, "final_distance": 1.505481243133545, "max_param_abs": 1.2752596139907837, "policy_loss": -0.0023780723726360975, "value_loss": 5.023116946220398}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:31:11.724478+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984671.7244933, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.967288471130305, "elapsed_seconds": 7082.238496069098, "env_steps_per_second": 10448.673825524615, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1595098760375923, "eta_seconds": 574.2355537353324, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005882952375078077, "chaser_grad_norm": 0.6506184786558151, "chaser_new_fields": 50.92499923706055, "chaser_partner_vision": 0.9362500309944153, "chaser_return": 16.10249900817871, "chaser_subspace_norm": 49.66584777832031, "collisions_per_episode": 14.975000381469727, "entropy": 0.19200139865279198, "explorer_grad_norm": 0.3906855434179306, "explorer_new_fields": 60.375, "explorer_partner_vision": 0.9362500309944153, "explorer_return": 28.05000114440918, "explorer_subspace_norm": 98.29364776611328, "final_distance": 2.046499013900757, "max_param_abs": 1.2868541479110718, "policy_loss": -0.004761775748342245, "value_loss": 5.3228806257247925}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:19.996128+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984859.9961395, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.01955163185724, "elapsed_seconds": 7270.385874686064, "env_steps_per_second": 10453.365379768331, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10629219115038105, "eta_seconds": 382.6518881413718, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0008754820732432833, "chaser_grad_norm": 0.3485313393175602, "chaser_new_fields": 51.29999923706055, "chaser_partner_vision": 0.9382500648498535, "chaser_return": 20.252500534057617, "chaser_subspace_norm": 54.76551818847656, "collisions_per_episode": 18.625, "entropy": 0.14456059038639069, "explorer_grad_norm": 0.26852596178650856, "explorer_new_fields": 57.275001525878906, "explorer_partner_vision": 0.9382500648498535, "explorer_return": 19.962499618530273, "explorer_subspace_norm": 96.9774398803711, "final_distance": 1.6334186792373657, "max_param_abs": 1.3034088611602783, "policy_loss": -0.003735009915661003, "value_loss": 5.188294887542725}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:28.143508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985048.1435242, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.073366262373068, "elapsed_seconds": 7464.118544543046, "env_steps_per_second": 10449.994803073585, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05316323749674534, "eta_seconds": 191.38765498828323, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0010697059647556961, "chaser_grad_norm": 0.7466687560081482, "chaser_new_fields": 48.900001525878906, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 19.55000114440918, "chaser_subspace_norm": 50.75027084350586, "collisions_per_episode": 18.80000114440918, "entropy": 0.16898205876350403, "explorer_grad_norm": 0.6775427758693695, "explorer_new_fields": 59.79999923706055, "explorer_partner_vision": 0.9455000162124634, "explorer_return": 23.17500114440918, "explorer_subspace_norm": 116.81356048583984, "final_distance": 1.6406723260879517, "max_param_abs": 1.3078556060791016, "policy_loss": -0.004742193066514577, "value_loss": 5.438898682594299}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:41.876178+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985241.8761919, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1234874252755738, "elapsed_seconds": 7644.554730992066, "env_steps_per_second": 10464.965300813285, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:42.312363+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985422.3123722, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.123490096754459, "elapsed_seconds": 7644.564348316053, "env_steps_per_second": 10464.952135254434, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0005562261293015069, "chaser_grad_norm": 0.6423081681132317, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 23.4325008392334, "chaser_subspace_norm": 57.10129928588867, "collisions_per_episode": 22.625, "entropy": 0.14692536741495132, "explorer_grad_norm": 0.5817638859152794, "explorer_new_fields": 52.775001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 12.71250057220459, "explorer_subspace_norm": 122.46495819091797, "final_distance": 1.791115403175354, "max_param_abs": 1.3355721235275269, "policy_loss": -0.004476253691127141, "value_loss": 5.1513988971710205}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:42.321981+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985422.3219912, "update": 20000} diff --git a/social/seed_0001/attempt_01/status.json b/social/seed_0001/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..104aa933d467c0a3a596adaeff0687ca997c9001 --- /dev/null +++ b/social/seed_0001/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.123490096754459, + "elapsed_seconds": 7644.564348316053, + "env_steps_per_second": 10464.952135254434, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0005562261293015069, + "chaser_grad_norm": 0.6423081681132317, + "chaser_new_fields": 46.20000076293945, + "chaser_partner_vision": 0.9297500252723694, + "chaser_return": 23.4325008392334, + "chaser_subspace_norm": 57.10129928588867, + "collisions_per_episode": 22.625, + "entropy": 0.14692536741495132, + "explorer_grad_norm": 0.5817638859152794, + "explorer_new_fields": 52.775001525878906, + "explorer_partner_vision": 0.9297500252723694, + "explorer_return": 12.71250057220459, + "explorer_subspace_norm": 122.46495819091797, + "final_distance": 1.791115403175354, + "max_param_abs": 1.3355721235275269, + "policy_loss": -0.004476253691127141, + "value_loss": 5.1513988971710205 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:42.321981+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985422.3219912, + "update": 20000 +} diff --git a/social/seed_0002/attempt_01/checkpoints/latest.safetensors b/social/seed_0002/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b8de8e354823b29ea046ddaad68749042efe2694 --- /dev/null +++ b/social/seed_0002/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:ddabd74e45c2c760a304a502299b61ff7db69c634e36f1f146feb9bf84e8fa5c +size 11460400 diff --git a/social/seed_0002/attempt_01/config.json b/social/seed_0002/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..02445b26fa34185cc9142980ac4d3fa93c021591 --- /dev/null +++ b/social/seed_0002/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0002.attempt_01", + "seed": 2, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0002/attempt_01/metrics.jsonl b/social/seed_0002/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..d005618474c8483f2a6a85d4c85c6725e2d934bc --- /dev/null +++ b/social/seed_0002/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.000559344379160191, "elapsed_seconds": 2.0136397649766877, "env_steps_per_second": 1986.452626518482, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 11.186328238824661, "eta_seconds": 40270.78165976878, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -6.514167283422356e-05, "chaser_grad_norm": 0.19650406017899513, "chaser_new_fields": 33.67499923706055, "chaser_partner_vision": 0.35100001096725464, "chaser_return": -1.357500433921814, "chaser_subspace_norm": 0.003058330388739705, "collisions_per_episode": 1.774999976158142, "entropy": 2.7724462151527405, "explorer_grad_norm": 0.07244767807424068, "explorer_new_fields": 36.95000076293945, "explorer_partner_vision": 0.35100001096725464, "explorer_return": 4.425000190734863, "explorer_subspace_norm": 0.0028640509117394686, "final_distance": 5.459163188934326, "max_param_abs": 0.15426841378211975, "policy_loss": -0.0008864622660913479, "value_loss": 3.1912264227867126}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.764320+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.7643328, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05428230598413696, "elapsed_seconds": 195.41630154289305, "env_steps_per_second": 10234.560700459313, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1170099333813415, "eta_seconds": 7621.23576017283, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0009260268793607196, "chaser_grad_norm": 0.3574873134493828, "chaser_new_fields": 54.125, "chaser_partner_vision": 0.9132500290870667, "chaser_return": 15.267501831054688, "chaser_subspace_norm": 17.622356414794922, "collisions_per_episode": 13.625, "entropy": 0.5445767194032669, "explorer_grad_norm": 0.2987947463989258, "explorer_new_fields": 61.875, "explorer_partner_vision": 0.9132500290870667, "explorer_return": 31.387500762939453, "explorer_subspace_norm": 9.840826034545898, "final_distance": 2.0231709480285645, "max_param_abs": 0.31068870425224304, "policy_loss": -0.0060679317787233344, "value_loss": 5.7330567836761475}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:33.166982+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977973.1669922, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10653719759248714, "elapsed_seconds": 383.5339113329537, "env_steps_per_second": 10429.325495881685, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0242067542572557, "eta_seconds": 7287.144315326121, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00029000148778379753, "chaser_grad_norm": 0.3401150330901146, "chaser_new_fields": 49.07500076293945, "chaser_partner_vision": 0.8642500638961792, "chaser_return": 13.402501106262207, "chaser_subspace_norm": 21.92249870300293, "collisions_per_episode": 12.875, "entropy": 0.43341150134801865, "explorer_grad_norm": 0.5098770186305046, "explorer_new_fields": 63.07500076293945, "explorer_partner_vision": 0.8642500638961792, "explorer_return": 33.375, "explorer_subspace_norm": 15.212387084960938, "final_distance": 2.160421133041382, "max_param_abs": 0.3274058997631073, "policy_loss": -0.006183066486382671, "value_loss": 5.659006714820862}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:41.284593+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978161.2846074, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15986839948137962, "elapsed_seconds": 575.5262381329667, "env_steps_per_second": 10425.241461560941, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.971710260270349, "eta_seconds": 7098.156936973256, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0002305257827490692, "chaser_grad_norm": 1.0451677739620209, "chaser_new_fields": 45.97500228881836, "chaser_partner_vision": 0.8577500581741333, "chaser_return": 18.255001068115234, "chaser_subspace_norm": 27.532257080078125, "collisions_per_episode": 17.75, "entropy": 0.32180481404066086, "explorer_grad_norm": 0.33585450798273087, "explorer_new_fields": 57.375, "explorer_partner_vision": 0.8577500581741333, "explorer_return": 23.36250114440918, "explorer_subspace_norm": 22.134403228759766, "final_distance": 2.2204418182373047, "max_param_abs": 0.34722819924354553, "policy_loss": -0.0028944216638393527, "value_loss": 6.222492933273315}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:53.276918+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978353.2769287, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21445828141693105, "elapsed_seconds": 772.0498131009517, "env_steps_per_second": 10362.025693481952, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9301245327523793, "eta_seconds": 6948.4483179085655, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0007190573187649818, "chaser_grad_norm": 0.4516584053635597, "chaser_new_fields": 51.17499923706055, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 17.110000610351562, "chaser_subspace_norm": 32.77616500854492, "collisions_per_episode": 16.0, "entropy": 0.20015956461429596, "explorer_grad_norm": 0.7182754725217819, "explorer_new_fields": 64.20000457763672, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 32.0, "explorer_subspace_norm": 28.8388671875, "final_distance": 1.7982981204986572, "max_param_abs": 0.3969484567642212, "policy_loss": -0.004851237590891344, "value_loss": 5.407433867454529}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:09.800493+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978549.8005064, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26813041656413894, "elapsed_seconds": 965.2694996309001, "env_steps_per_second": 10359.801075061216, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8769129159489724, "eta_seconds": 6756.886497416301, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00036652923423829975, "chaser_grad_norm": 0.5683717727661133, "chaser_new_fields": 49.375, "chaser_partner_vision": 0.9327500462532043, "chaser_return": 18.80500030517578, "chaser_subspace_norm": 32.6031494140625, "collisions_per_episode": 17.950000762939453, "entropy": 0.270922914147377, "explorer_grad_norm": 0.6707459539175034, "explorer_new_fields": 63.54999923706055, "explorer_partner_vision": 0.9327500462532043, "explorer_return": 29.262500762939453, "explorer_subspace_norm": 27.317832946777344, "final_distance": 1.7081447839736938, "max_param_abs": 0.41597047448158264, "policy_loss": -0.006548921366217542, "value_loss": 5.810567140579224}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:23.020181+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978743.0201955, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3211669837827665, "elapsed_seconds": 1156.2011416179594, "env_steps_per_second": 10378.816944608354, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8199462414356768, "eta_seconds": 6551.806469168437, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005399066968909683, "chaser_grad_norm": 0.35859810560941696, "chaser_new_fields": 50.67499923706055, "chaser_partner_vision": 0.9197500348091125, "chaser_return": 17.277502059936523, "chaser_subspace_norm": 41.36969757080078, "collisions_per_episode": 16.075000762939453, "entropy": 0.2049768827855587, "explorer_grad_norm": 0.34768597036600113, "explorer_new_fields": 60.07500076293945, "explorer_partner_vision": 0.9197500348091125, "explorer_return": 27.649999618530273, "explorer_subspace_norm": 35.32948303222656, "final_distance": 1.9193508625030518, "max_param_abs": 0.48432809114456177, "policy_loss": -0.002992647288138528, "value_loss": 5.428441405296326}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:33.951823+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978933.9518385, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.37317790367304243, "elapsed_seconds": 1343.4404532229528, "env_steps_per_second": 10421.00523801676, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7592672601729142, "eta_seconds": 6333.362136622492, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0011589826203097786, "chaser_grad_norm": 0.4446157030761242, "chaser_new_fields": 51.17499923706055, "chaser_partner_vision": 0.9312500357627869, "chaser_return": 18.310001373291016, "chaser_subspace_norm": 38.99213409423828, "collisions_per_episode": 17.30000114440918, "entropy": 0.17442743480205536, "explorer_grad_norm": 0.2922973409295082, "explorer_new_fields": 63.04999923706055, "explorer_partner_vision": 0.9312500357627869, "explorer_return": 30.0, "explorer_subspace_norm": 34.40361404418945, "final_distance": 1.7453426122665405, "max_param_abs": 0.5436623692512512, "policy_loss": -0.005367213411254568, "value_loss": 4.935069918632507}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:41.191133+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979121.1911426, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.42792468223138713, "elapsed_seconds": 1540.5288560329936, "env_steps_per_second": 10386.04368710204, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7116987289255485, "eta_seconds": 6162.115424131975, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0012170000431417707, "chaser_grad_norm": 0.6011382415890694, "chaser_new_fields": 50.625, "chaser_partner_vision": 0.9262500405311584, "chaser_return": 15.64000129699707, "chaser_subspace_norm": 36.502601623535156, "collisions_per_episode": 14.699999809265137, "entropy": 0.2008986994624138, "explorer_grad_norm": 0.42237236350774765, "explorer_new_fields": 61.275001525878906, "explorer_partner_vision": 0.9262500405311584, "explorer_return": 28.75, "explorer_subspace_norm": 36.72615051269531, "final_distance": 1.90142822265625, "max_param_abs": 0.6131850481033325, "policy_loss": -0.006863729195669244, "value_loss": 5.420848727226257}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:58.279537+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979318.2795494, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4799456335416633, "elapsed_seconds": 1727.8042807499878, "env_steps_per_second": 10417.846627968038, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.653146071087951, "eta_seconds": 5951.325855916624, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0008311042595066098, "chaser_grad_norm": 0.2264445275068283, "chaser_new_fields": 49.95000076293945, "chaser_partner_vision": 0.95250004529953, "chaser_return": 17.14750099182129, "chaser_subspace_norm": 35.987003326416016, "collisions_per_episode": 16.325000762939453, "entropy": 0.16825879365205765, "explorer_grad_norm": 0.40973786264657974, "explorer_new_fields": 62.25, "explorer_partner_vision": 0.95250004529953, "explorer_return": 28.98750114440918, "explorer_subspace_norm": 41.855918884277344, "final_distance": 1.7524757385253906, "max_param_abs": 0.6447548270225525, "policy_loss": -0.004449624171947164, "value_loss": 5.476781845092773}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:05:05.554959+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979505.55497, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.53217857192859, "elapsed_seconds": 1915.842858942924, "env_steps_per_second": 10439.269539587971, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.59653571578577, "eta_seconds": 5747.528576828772, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0002437988094903254, "chaser_grad_norm": 1.4535012543201447, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9200000166893005, "chaser_return": 16.39000129699707, "chaser_subspace_norm": 37.28496551513672, "collisions_per_episode": 15.65000057220459, "entropy": 0.18439680710434914, "explorer_grad_norm": 0.29821958020329475, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.9200000166893005, "explorer_return": 29.4375, "explorer_subspace_norm": 36.55598449707031, "final_distance": 1.9152134656906128, "max_param_abs": 0.6670604348182678, "policy_loss": -0.0038811418963593525, "value_loss": 5.663198471069336}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:13.593539+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979693.5935533, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5841588624880468, "elapsed_seconds": 2102.9719049569685, "env_steps_per_second": 10461.38559823041, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5400551829230325, "eta_seconds": 5544.198658522917, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0003493183117533327, "chaser_grad_norm": 1.0417082011699677, "chaser_new_fields": 49.42499923706055, "chaser_partner_vision": 0.940000057220459, "chaser_return": 20.032501220703125, "chaser_subspace_norm": 40.54246139526367, "collisions_per_episode": 18.975000381469727, "entropy": 0.13690808042883873, "explorer_grad_norm": 0.40253305435180664, "explorer_new_fields": 61.72500228881836, "explorer_partner_vision": 0.940000057220459, "explorer_return": 26.162500381469727, "explorer_subspace_norm": 50.14556121826172, "final_distance": 1.6318782567977905, "max_param_abs": 0.6223595142364502, "policy_loss": -0.004064980964656684, "value_loss": 5.521803498268127}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:20.722585+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979880.7225952, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.636719823147222, "elapsed_seconds": 2292.191363329999, "env_steps_per_second": 10470.330001215, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4856795873435178, "eta_seconds": 5348.446514436664, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0012642774045417138, "chaser_grad_norm": 1.470617026090622, "chaser_new_fields": 51.75, "chaser_partner_vision": 0.9350000619888306, "chaser_return": 19.907501220703125, "chaser_subspace_norm": 40.65605545043945, "collisions_per_episode": 18.524999618530273, "entropy": 0.16025221347808838, "explorer_grad_norm": 0.4301885664463043, "explorer_new_fields": 62.025001525878906, "explorer_partner_vision": 0.9350000619888306, "explorer_return": 28.375, "explorer_subspace_norm": 52.60136413574219, "final_distance": 1.7828577756881714, "max_param_abs": 0.6287584900856018, "policy_loss": -0.004355883705429164, "value_loss": 5.4184465408325195}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:29.942044+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980069.9420583, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6903586985955351, "elapsed_seconds": 2485.2913149439264, "env_steps_per_second": 10461.55025918425, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.43382191246765, "eta_seconds": 5161.758884883539, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006640758652274314, "chaser_grad_norm": 1.1443726420402527, "chaser_new_fields": 52.04999923706055, "chaser_partner_vision": 0.9587500691413879, "chaser_return": 20.61250114440918, "chaser_subspace_norm": 43.30437088012695, "collisions_per_episode": 18.975000381469727, "entropy": 0.1459057554602623, "explorer_grad_norm": 0.5141990929841995, "explorer_new_fields": 62.82500076293945, "explorer_partner_vision": 0.9587500691413879, "explorer_return": 28.5625, "explorer_subspace_norm": 50.86378479003906, "final_distance": 1.767652153968811, "max_param_abs": 0.6292352676391602, "policy_loss": -0.004648410504216827, "value_loss": 5.486109972000122}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:43.041995+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980263.042008, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7451293944622093, "elapsed_seconds": 2682.4658200639533, "env_steps_per_second": 10438.157232263427, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3838117325726744, "eta_seconds": 4981.722237261628, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00015850291245733672, "chaser_grad_norm": 0.33187608048319817, "chaser_new_fields": 52.650001525878906, "chaser_partner_vision": 0.9392500519752502, "chaser_return": 20.150001525878906, "chaser_subspace_norm": 40.8137092590332, "collisions_per_episode": 18.450000762939453, "entropy": 0.15192924439907074, "explorer_grad_norm": 0.2818678580224514, "explorer_new_fields": 60.42499923706055, "explorer_partner_vision": 0.9392500519752502, "explorer_return": 25.375, "explorer_subspace_norm": 54.215301513671875, "final_distance": 1.606723427772522, "max_param_abs": 0.6400149464607239, "policy_loss": -0.0036150371362566602, "value_loss": 5.537940144538879}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:21:00.216501+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980460.2165146, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7970108866988772, "elapsed_seconds": 2869.239192115958, "env_steps_per_second": 10455.73338132054, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3283514778314618, "eta_seconds": 4782.065320193263, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0004852490230413675, "chaser_grad_norm": 0.24940641596913338, "chaser_new_fields": 52.72500228881836, "chaser_partner_vision": 0.9097500443458557, "chaser_return": 19.13250160217285, "chaser_subspace_norm": 40.0906982421875, "collisions_per_episode": 17.625, "entropy": 0.1572512723505497, "explorer_grad_norm": 0.31524574756622314, "explorer_new_fields": 60.32500076293945, "explorer_partner_vision": 0.9097500443458557, "explorer_return": 25.86250114440918, "explorer_subspace_norm": 53.62427520751953, "final_distance": 1.9093776941299438, "max_param_abs": 0.7037089467048645, "policy_loss": -0.0046444286094065035, "value_loss": 5.519266963005066}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:06.989871+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980646.989883, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8489032179863878, "elapsed_seconds": 3056.051584750996, "env_steps_per_second": 10471.02743935107, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2733548269795816, "eta_seconds": 4584.077377126494, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0008618371887227916, "chaser_grad_norm": 0.2730023190379143, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.905250072479248, "chaser_return": 20.100000381469727, "chaser_subspace_norm": 41.14104080200195, "collisions_per_episode": 19.0, "entropy": 0.14656323939561844, "explorer_grad_norm": 1.0630464255809784, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.905250072479248, "explorer_return": 27.649999618530273, "explorer_subspace_norm": 58.15238952636719, "final_distance": 1.720741629600525, "max_param_abs": 0.7192366719245911, "policy_loss": -0.0037242480302239755, "value_loss": 5.313792109489441}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:13.802266+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980833.8022804, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9020182271269409, "elapsed_seconds": 3247.265617656987, "env_steps_per_second": 10470.347671938263, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2203776014070378, "eta_seconds": 4393.359365065336, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.002035531687707204, "chaser_grad_norm": 0.3810575194656849, "chaser_new_fields": 50.625, "chaser_partner_vision": 0.9237500429153442, "chaser_return": 17.89000129699707, "chaser_subspace_norm": 39.20629119873047, "collisions_per_episode": 16.899999618530273, "entropy": 0.1614399030804634, "explorer_grad_norm": 0.2919735386967659, "explorer_new_fields": 62.42499923706055, "explorer_partner_vision": 0.9237500429153442, "explorer_return": 29.1875, "explorer_subspace_norm": 59.44832992553711, "final_distance": 1.6947342157363892, "max_param_abs": 0.725928008556366, "policy_loss": -0.005521501743235291, "value_loss": 5.260424375534058}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:25.016297+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981025.0163078, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9578408143658048, "elapsed_seconds": 3448.2269317168975, "env_steps_per_second": 10440.14814363605, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.170694328669317, "eta_seconds": 4214.499583209541, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00018704964833737847, "chaser_grad_norm": 0.5172083638608456, "chaser_new_fields": 48.375, "chaser_partner_vision": 0.8885000348091125, "chaser_return": 19.005001068115234, "chaser_subspace_norm": 39.25627517700195, "collisions_per_episode": 18.25, "entropy": 0.17996235564351082, "explorer_grad_norm": 0.46196916699409485, "explorer_new_fields": 60.97500228881836, "explorer_partner_vision": 0.8885000348091125, "explorer_return": 25.625, "explorer_subspace_norm": 60.8056640625, "final_distance": 1.7774757146835327, "max_param_abs": 0.7244802117347717, "policy_loss": -0.0047884098817583265, "value_loss": 5.587342739105225}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:45.977613+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981225.9776285, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0116620052997152, "elapsed_seconds": 3641.983219078975, "env_steps_per_second": 10433.875642515965, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1181527426996853, "eta_seconds": 4025.349873718867, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0010147388697726994, "chaser_grad_norm": 0.5544580444693565, "chaser_new_fields": 48.275001525878906, "chaser_partner_vision": 0.9167500138282776, "chaser_return": 21.452499389648438, "chaser_subspace_norm": 41.85272216796875, "collisions_per_episode": 20.475000381469727, "entropy": 0.12107830867171288, "explorer_grad_norm": 0.6385203264653683, "explorer_new_fields": 65.75, "explorer_partner_vision": 0.9167500138282776, "explorer_return": 30.625, "explorer_subspace_norm": 65.23358917236328, "final_distance": 1.7693363428115845, "max_param_abs": 0.7554123997688293, "policy_loss": -0.004221328095534105, "value_loss": 4.956096410751343}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:59.733900+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981419.7339149, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0637606233272041, "elapsed_seconds": 3829.538243977935, "env_steps_per_second": 10445.123524461784, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0637606233272041, "eta_seconds": 3829.538243977935, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0008771836584262616, "chaser_grad_norm": 0.3000178560614586, "chaser_new_fields": 49.45000076293945, "chaser_partner_vision": 0.940500020980835, "chaser_return": 19.53000259399414, "chaser_subspace_norm": 38.11256408691406, "collisions_per_episode": 18.600000381469727, "entropy": 0.17459405213594437, "explorer_grad_norm": 0.23574042320251465, "explorer_new_fields": 61.525001525878906, "explorer_partner_vision": 0.940500020980835, "explorer_return": 26.387500762939453, "explorer_subspace_norm": 67.67754364013672, "final_distance": 1.7764631509780884, "max_param_abs": 0.7584251761436462, "policy_loss": -0.00476532314065814, "value_loss": 5.054116249084473}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:40:07.288925+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981607.2889402, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1181557890819385, "elapsed_seconds": 4025.360840694979, "env_steps_per_second": 10433.847215731521, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0116647615503254, "eta_seconds": 3641.9931415811716, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0008830732321150059, "chaser_grad_norm": 0.33826936036348343, "chaser_new_fields": 44.150001525878906, "chaser_partner_vision": 0.9262500405311584, "chaser_return": 21.32000160217285, "chaser_subspace_norm": 38.71753692626953, "collisions_per_episode": 21.200000762939453, "entropy": 0.19522231817245483, "explorer_grad_norm": 0.40368204563856125, "explorer_new_fields": 60.22500228881836, "explorer_partner_vision": 0.9262500405311584, "explorer_return": 23.137500762939453, "explorer_subspace_norm": 68.68656921386719, "final_distance": 1.7186778783798218, "max_param_abs": 0.7805008292198181, "policy_loss": -0.006005165216512509, "value_loss": 5.52934718132019}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:23.111521+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981803.111534, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1704574714274754, "elapsed_seconds": 4213.646897138911, "env_steps_per_second": 10442.26084294729, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9576470220770253, "eta_seconds": 3447.5292794772913, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": -0.00017284608762668796, "chaser_grad_norm": 0.380210243165493, "chaser_new_fields": 47.72500228881836, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 21.78999900817871, "chaser_subspace_norm": 42.53746795654297, "collisions_per_episode": 21.0, "entropy": 0.1875956691801548, "explorer_grad_norm": 0.5802170857787132, "explorer_new_fields": 63.29999923706055, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 25.787500381469727, "explorer_subspace_norm": 65.27233123779297, "final_distance": 1.7346343994140625, "max_param_abs": 0.8176241517066956, "policy_loss": -0.005598483552692102, "value_loss": 5.578966021537781}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:31.397578+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981991.3975918, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2245746260458126, "elapsed_seconds": 4408.468653764925, "env_steps_per_second": 10434.462307158527, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9051203757729918, "eta_seconds": 3258.4333527827703, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0009305918498245802, "chaser_grad_norm": 0.3717942014336586, "chaser_new_fields": 53.22500228881836, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 22.332502365112305, "chaser_subspace_norm": 42.238121032714844, "collisions_per_episode": 20.475000381469727, "entropy": 0.16506826505064964, "explorer_grad_norm": 0.3610350862145424, "explorer_new_fields": 64.9749984741211, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 28.600000381469727, "explorer_subspace_norm": 70.4192886352539, "final_distance": 1.7364791631698608, "max_param_abs": 0.816246747970581, "policy_loss": -0.005369213365952419, "value_loss": 4.89491331577301}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:46.219333+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982186.2193432, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2771941950160868, "elapsed_seconds": 4597.899102057912, "env_steps_per_second": 10439.550528308966, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8514627966773912, "eta_seconds": 3065.2660680386084, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00047102825637890255, "chaser_grad_norm": 0.6337175667285919, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9145000576972961, "chaser_return": 20.697500228881836, "chaser_subspace_norm": 42.40870666503906, "collisions_per_episode": 19.625, "entropy": 0.1541372761130333, "explorer_grad_norm": 0.6014879569411278, "explorer_new_fields": 58.0, "explorer_partner_vision": 0.9145000576972961, "explorer_return": 20.625, "explorer_subspace_norm": 97.36343383789062, "final_distance": 1.684302568435669, "max_param_abs": 0.8829600811004639, "policy_loss": -0.0030470068419852225, "value_loss": 5.799917936325073}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:55.649782+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982375.6497972, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3318606349472086, "elapsed_seconds": 4794.698285809951, "env_steps_per_second": 10428.184844910149, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7991163809683252, "eta_seconds": 2876.8189714859705, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0015040896390228209, "chaser_grad_norm": 0.31950483471155167, "chaser_new_fields": 49.79999923706055, "chaser_partner_vision": 0.9575000405311584, "chaser_return": 20.77250099182129, "chaser_subspace_norm": 47.42896270751953, "collisions_per_episode": 19.575000762939453, "entropy": 0.13806582614779472, "explorer_grad_norm": 0.21729058027267456, "explorer_new_fields": 58.875, "explorer_partner_vision": 0.9575000405311584, "explorer_return": 21.887500762939453, "explorer_subspace_norm": 85.16802215576172, "final_distance": 1.4235281944274902, "max_param_abs": 0.9919006824493408, "policy_loss": -0.0036901400499047554, "value_loss": 5.128406882286072}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:56:12.448966+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982572.4489808, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3853415241924814, "elapsed_seconds": 4987.229487092933, "env_steps_per_second": 10426.630684346334, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7459531284113362, "eta_seconds": 2685.43126228081, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0006272515373826337, "chaser_grad_norm": 0.22158554196357727, "chaser_new_fields": 53.04999923706055, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 22.587499618530273, "chaser_subspace_norm": 46.47908020019531, "collisions_per_episode": 20.92500114440918, "entropy": 0.11835101991891861, "explorer_grad_norm": 0.4718587026000023, "explorer_new_fields": 61.82500076293945, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 23.6875, "explorer_subspace_norm": 81.56110382080078, "final_distance": 1.7054131031036377, "max_param_abs": 0.9810304045677185, "policy_loss": -0.00362448597699494, "value_loss": 5.010036945343018}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:24.980169+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982764.980184, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4376823706416568, "elapsed_seconds": 5175.656534309965, "env_steps_per_second": 10433.458951927816, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6922174377163534, "eta_seconds": 2491.982775778872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -0.00035921554941475975, "chaser_grad_norm": 0.2840604856610298, "chaser_new_fields": 47.875, "chaser_partner_vision": 0.9335000514984131, "chaser_return": 21.565000534057617, "chaser_subspace_norm": 43.226806640625, "collisions_per_episode": 20.80000114440918, "entropy": 0.1622718647122383, "explorer_grad_norm": 0.5628661960363388, "explorer_new_fields": 58.92499923706055, "explorer_partner_vision": 0.9335000514984131, "explorer_return": 21.23750114440918, "explorer_subspace_norm": 85.16036987304688, "final_distance": 1.7389236688613892, "max_param_abs": 0.9930444359779358, "policy_loss": -0.005370300269220252, "value_loss": 5.5293720960617065}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:33.407213+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982953.407225, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4905733096119365, "elapsed_seconds": 5366.063914602972, "env_steps_per_second": 10435.954713026069, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.63881713269083, "eta_seconds": 2299.7416776869877, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0012381750285173077, "chaser_grad_norm": 0.6503573581576347, "chaser_new_fields": 48.45000076293945, "chaser_partner_vision": 0.9425000548362732, "chaser_return": 21.477500915527344, "chaser_subspace_norm": 42.8387451171875, "collisions_per_episode": 20.575000762939453, "entropy": 0.1910560093820095, "explorer_grad_norm": 0.24509960785508156, "explorer_new_fields": 59.79999923706055, "explorer_partner_vision": 0.9425000548362732, "explorer_return": 22.4375, "explorer_subspace_norm": 83.75753784179688, "final_distance": 1.8205242156982422, "max_param_abs": 1.0028620958328247, "policy_loss": -0.005980243036518118, "value_loss": 5.329637408256531}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:43.814595+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983143.8146086, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5443565377388666, "elapsed_seconds": 5559.68353585992, "env_steps_per_second": 10432.248459089515, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5857904108664668, "eta_seconds": 2108.8454791192803, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00016042733250598573, "chaser_grad_norm": 0.5359846204519272, "chaser_new_fields": 49.42499923706055, "chaser_partner_vision": 0.9392500519752502, "chaser_return": 19.6875, "chaser_subspace_norm": 45.21609878540039, "collisions_per_episode": 18.67500114440918, "entropy": 0.14442123845219612, "explorer_grad_norm": 0.32353125512599945, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9392500519752502, "explorer_return": 20.387500762939453, "explorer_subspace_norm": 78.26863098144531, "final_distance": 1.8166168928146362, "max_param_abs": 1.0371679067611694, "policy_loss": -0.0042871495517893266, "value_loss": 4.297364234924316}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:57.434217+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983337.43423, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5986157729966606, "elapsed_seconds": 5755.016782787978, "env_steps_per_second": 10425.686364538005, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5328719243322202, "eta_seconds": 1918.3389275959926, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0008282863717159472, "chaser_grad_norm": 0.6104339882731438, "chaser_new_fields": 52.29999923706055, "chaser_partner_vision": 0.9232500195503235, "chaser_return": 19.227500915527344, "chaser_subspace_norm": 46.33172607421875, "collisions_per_episode": 17.774999618530273, "entropy": 0.17603350430727005, "explorer_grad_norm": 0.28511451929807663, "explorer_new_fields": 61.75, "explorer_partner_vision": 0.9232500195503235, "explorer_return": 27.587499618530273, "explorer_subspace_norm": 74.4327163696289, "final_distance": 1.9469364881515503, "max_param_abs": 1.0185006856918335, "policy_loss": -0.005088821679354538, "value_loss": 5.128989219665527}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:12:12.767463+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983532.7674768, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.650965401637756, "elapsed_seconds": 5943.475445895921, "env_steps_per_second": 10431.606988939802, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4793125359593485, "eta_seconds": 1725.5251294536545, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004822085617821603, "chaser_grad_norm": 0.27475081756711006, "chaser_new_fields": 50.47500228881836, "chaser_partner_vision": 0.9410000443458557, "chaser_return": 20.09000015258789, "chaser_subspace_norm": 47.13646697998047, "collisions_per_episode": 18.850000381469727, "entropy": 0.1790294386446476, "explorer_grad_norm": 0.4796280637383461, "explorer_new_fields": 61.45000076293945, "explorer_partner_vision": 0.9410000443458557, "explorer_return": 26.48750114440918, "explorer_subspace_norm": 86.21294403076172, "final_distance": 1.7954254150390625, "max_param_abs": 0.9941220283508301, "policy_loss": -0.004334686720746106, "value_loss": 5.276227235794067}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:15:21.226125+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983721.226136, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.703834482254428, "elapsed_seconds": 6133.804136115941, "env_steps_per_second": 10433.981682454927, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.425958620563607, "eta_seconds": 1533.4510340289853, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0008224090597312994, "chaser_grad_norm": 0.4192977398633957, "chaser_new_fields": 49.375, "chaser_partner_vision": 0.9282500147819519, "chaser_return": 17.44500160217285, "chaser_subspace_norm": 50.22223663330078, "collisions_per_episode": 16.600000381469727, "entropy": 0.1626329980790615, "explorer_grad_norm": 0.48433154821395874, "explorer_new_fields": 60.375, "explorer_partner_vision": 0.9282500147819519, "explorer_return": 27.200000762939453, "explorer_subspace_norm": 85.01496887207031, "final_distance": 1.5725704431533813, "max_param_abs": 1.0425910949707031, "policy_loss": -0.004472904670983269, "value_loss": 4.995472550392151}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:31.554817+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983911.5548296, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7557664958919244, "elapsed_seconds": 6320.759385210928, "env_steps_per_second": 10441.783332936908, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3724353173104082, "eta_seconds": 1340.7671423174695, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00024244602807144133, "chaser_grad_norm": 0.3281093128025532, "chaser_new_fields": 50.79999923706055, "chaser_partner_vision": 0.952250063419342, "chaser_return": 21.552501678466797, "chaser_subspace_norm": 54.64078903198242, "collisions_per_episode": 20.375, "entropy": 0.11049539595842361, "explorer_grad_norm": 0.4228176400065422, "explorer_new_fields": 57.97500228881836, "explorer_partner_vision": 0.952250063419342, "explorer_return": 18.1875, "explorer_subspace_norm": 84.75565338134766, "final_distance": 1.5323259830474854, "max_param_abs": 1.0332835912704468, "policy_loss": -0.0026226594373603973, "value_loss": 4.9081069231033325}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:38.510064+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984098.5100758, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8080765154305523, "elapsed_seconds": 6509.075455549988, "env_steps_per_second": 10446.952176905485, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3190723262524504, "eta_seconds": 1148.6603745088214, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00035569935515211704, "chaser_grad_norm": 0.41620323807001114, "chaser_new_fields": 45.97500228881836, "chaser_partner_vision": 0.9177500605583191, "chaser_return": 32.67250442504883, "chaser_subspace_norm": 55.55110168457031, "collisions_per_episode": 31.125, "entropy": 0.16144365072250366, "explorer_grad_norm": 0.2157963328063488, "explorer_new_fields": 53.35000228881836, "explorer_partner_vision": 0.9177500605583191, "explorer_return": 9.58750057220459, "explorer_subspace_norm": 98.13042449951172, "final_distance": 1.5235016345977783, "max_param_abs": 1.052293300628662, "policy_loss": -0.003959131435957053, "value_loss": 6.142445683479309}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:46.826136+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984286.8261507, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8600403105213708, "elapsed_seconds": 6696.145117876935, "env_steps_per_second": 10453.775831876543, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2657200443601958, "eta_seconds": 956.5921596967049, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0006885582464395412, "chaser_grad_norm": 0.5522056818008423, "chaser_new_fields": 52.57500076293945, "chaser_partner_vision": 0.9415000677108765, "chaser_return": 22.197500228881836, "chaser_subspace_norm": 51.88753128051758, "collisions_per_episode": 20.67500114440918, "entropy": 0.14425231888890266, "explorer_grad_norm": 0.26566508039832115, "explorer_new_fields": 59.95000076293945, "explorer_partner_vision": 0.9415000677108765, "explorer_return": 20.662500381469727, "explorer_subspace_norm": 96.8212661743164, "final_distance": 1.5800882577896118, "max_param_abs": 1.0698024034500122, "policy_loss": -0.004325457669796862, "value_loss": 5.417537450790405}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:53.895797+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984473.8958085, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.911986651231095, "elapsed_seconds": 6883.151944431942, "env_steps_per_second": 10460.324075548513, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21244296124789944, "eta_seconds": 764.794660492438, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00019948413172876656, "chaser_grad_norm": 0.4867020361125469, "chaser_new_fields": 51.79999923706055, "chaser_partner_vision": 0.9305000305175781, "chaser_return": 18.61250114440918, "chaser_subspace_norm": 48.81812286376953, "collisions_per_episode": 17.375, "entropy": 0.17995086684823036, "explorer_grad_norm": 0.6950601935386658, "explorer_new_fields": 58.625, "explorer_partner_vision": 0.9305000305175781, "explorer_return": 22.950000762939453, "explorer_subspace_norm": 108.49473571777344, "final_distance": 1.6940453052520752, "max_param_abs": 1.097337245941162, "policy_loss": -0.005204732599665363, "value_loss": 5.352016091346741}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:31:00.902624+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984660.9026384, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9645786981997015, "elapsed_seconds": 7072.483313518926, "env_steps_per_second": 10463.085838399975, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1592901647188947, "eta_seconds": 573.444592988021, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0011511704218403196, "chaser_grad_norm": 0.19960980489850044, "chaser_new_fields": 50.10000228881836, "chaser_partner_vision": 0.9410000443458557, "chaser_return": 21.48750114440918, "chaser_subspace_norm": 50.46564483642578, "collisions_per_episode": 20.325000762939453, "entropy": 0.11301277205348015, "explorer_grad_norm": 0.29485826566815376, "explorer_new_fields": 57.29999923706055, "explorer_partner_vision": 0.9410000443458557, "explorer_return": 18.625, "explorer_subspace_norm": 114.05460357666016, "final_distance": 1.7135066986083984, "max_param_abs": 1.111340880393982, "policy_loss": -0.002453547725455252, "value_loss": 5.45446515083313}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:10.233992+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984850.234004, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0169435662946973, "elapsed_seconds": 7260.996838660911, "env_steps_per_second": 10466.882397653832, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10615492454182618, "eta_seconds": 382.15772835057425, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0013548575220753145, "chaser_grad_norm": 0.49625610560178757, "chaser_new_fields": 52.04999923706055, "chaser_partner_vision": 0.9305000305175781, "chaser_return": 17.35500144958496, "chaser_subspace_norm": 46.10704803466797, "collisions_per_episode": 16.100000381469727, "entropy": 0.16917210444808006, "explorer_grad_norm": 0.6932548731565475, "explorer_new_fields": 59.275001525878906, "explorer_partner_vision": 0.9305000305175781, "explorer_return": 24.262500762939453, "explorer_subspace_norm": 97.49002075195312, "final_distance": 1.591600775718689, "max_param_abs": 1.0971969366073608, "policy_loss": -0.004456889094319294, "value_loss": 5.270473003387451}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:18.747518+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985038.7475302, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.0714483991033115, "elapsed_seconds": 7457.214236771921, "env_steps_per_second": 10459.66999518102, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.053114061515469525, "eta_seconds": 191.2106214556903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00027281455671701593, "chaser_grad_norm": 0.36930448189377785, "chaser_new_fields": 46.525001525878906, "chaser_partner_vision": 0.9000000357627869, "chaser_return": 20.617502212524414, "chaser_subspace_norm": 48.53520584106445, "collisions_per_episode": 20.075000762939453, "entropy": 0.18093863502144814, "explorer_grad_norm": 0.42068637907505035, "explorer_new_fields": 55.650001525878906, "explorer_partner_vision": 0.9000000357627869, "explorer_return": 17.92500114440918, "explorer_subspace_norm": 92.5508804321289, "final_distance": 1.8830817937850952, "max_param_abs": 1.115954875946045, "policy_loss": -0.005136819275563198, "value_loss": 5.5276947021484375}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:34.964916+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985234.964929, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.122354070891937, "elapsed_seconds": 7640.474655210972, "env_steps_per_second": 10470.553677635491, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:38.225335+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985418.2253711, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.122357136064416, "elapsed_seconds": 7640.485689831898, "env_steps_per_second": 10470.538555744632, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0014266868641836794, "chaser_grad_norm": 0.20290611311793327, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.48750114440918, "chaser_subspace_norm": 48.7114143371582, "collisions_per_episode": 17.625, "entropy": 0.18204283714294434, "explorer_grad_norm": 0.6118505597114563, "explorer_new_fields": 60.95000076293945, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 26.837499618530273, "explorer_subspace_norm": 104.41957092285156, "final_distance": 1.6572967767715454, "max_param_abs": 1.1167881488800049, "policy_loss": -0.005259661337355226, "value_loss": 5.285007953643799}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:38.236370+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985418.2363846, "update": 20000} diff --git a/social/seed_0002/attempt_01/status.json b/social/seed_0002/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..b8f049c759a709c14e89e4257be3c120df3727e1 --- /dev/null +++ b/social/seed_0002/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.122357136064416, + "elapsed_seconds": 7640.485689831898, + "env_steps_per_second": 10470.538555744632, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0014266868641836794, + "chaser_grad_norm": 0.20290611311793327, + "chaser_new_fields": 49.29999923706055, + "chaser_partner_vision": 0.9287500381469727, + "chaser_return": 18.48750114440918, + "chaser_subspace_norm": 48.7114143371582, + "collisions_per_episode": 17.625, + "entropy": 0.18204283714294434, + "explorer_grad_norm": 0.6118505597114563, + "explorer_new_fields": 60.95000076293945, + "explorer_partner_vision": 0.9287500381469727, + "explorer_return": 26.837499618530273, + "explorer_subspace_norm": 104.41957092285156, + "final_distance": 1.6572967767715454, + "max_param_abs": 1.1167881488800049, + "policy_loss": -0.005259661337355226, + "value_loss": 5.285007953643799 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:38.236370+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985418.2363846, + "update": 20000 +} diff --git a/social/seed_0003/attempt_01/checkpoints/latest.safetensors b/social/seed_0003/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7901821a7eb43afdbb9febd0e645f43b44c87534 --- /dev/null +++ b/social/seed_0003/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:77872aaef8ab25b68bf6d92a4af4bf74e5cc79b997d040bc34175db56af8e630 +size 11460400 diff --git a/social/seed_0003/attempt_01/config.json b/social/seed_0003/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..9d4f53f05dc488959bfbd2b1ae56f50829353d20 --- /dev/null +++ b/social/seed_0003/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0003.attempt_01", + "seed": 3, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0003/attempt_01/metrics.jsonl b/social/seed_0003/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..6b8b84533f88aaf210aa56f789832dbd29d93c2c --- /dev/null +++ b/social/seed_0003/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005114572672432081, "elapsed_seconds": 1.8412461620755494, "env_steps_per_second": 2172.441731251725, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.22863388759692, "eta_seconds": 36823.08199534891, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 2.109088497995182e-05, "chaser_grad_norm": 0.197127316147089, "chaser_new_fields": 31.350000381469727, "chaser_partner_vision": 0.33500000834465027, "chaser_return": -1.325000286102295, "chaser_subspace_norm": 0.0024993540719151497, "collisions_per_episode": 2.25, "entropy": 2.7725199460983276, "explorer_grad_norm": 0.1339152418076992, "explorer_new_fields": 33.22500228881836, "explorer_partner_vision": 0.33500000834465027, "explorer_return": -1.625, "explorer_subspace_norm": 0.0022529521957039833, "final_distance": 4.660165309906006, "max_param_abs": 0.15429139137268066, "policy_loss": -0.0007672591665763839, "value_loss": 3.364598035812378}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:19.809327+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977779.8093414, "update": 1} +{"device": "cuda", "elapsed_hours": 0.053628648269449415, "elapsed_seconds": 193.06313377001788, "env_steps_per_second": 10359.305585407388, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.091517282508527, "eta_seconds": 7529.4622170306975, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0010924493536723645, "chaser_grad_norm": 0.28148502111434937, "chaser_new_fields": 57.70000076293945, "chaser_partner_vision": 0.4857500195503235, "chaser_return": 9.8149995803833, "chaser_subspace_norm": 14.215556144714355, "collisions_per_episode": 7.75, "entropy": 0.6814913600683212, "explorer_grad_norm": 0.515291765332222, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.4857500195503235, "explorer_return": 39.61249923706055, "explorer_subspace_norm": 6.770150184631348, "final_distance": 4.3212571144104, "max_param_abs": 0.3189601004123688, "policy_loss": -0.005717624283398948, "value_loss": 5.426798343658447}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:31.031217+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977971.0312307, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10677352766917061, "elapsed_seconds": 384.3846996090142, "env_steps_per_second": 10406.24146608513, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0286970257142416, "eta_seconds": 7303.309292571269, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0009282158441337174, "chaser_grad_norm": 0.3383422791957855, "chaser_new_fields": 50.70000076293945, "chaser_partner_vision": 0.627500057220459, "chaser_return": 9.902501106262207, "chaser_subspace_norm": 26.45131492614746, "collisions_per_episode": 9.125, "entropy": 0.5642000138759613, "explorer_grad_norm": 0.2845703363418579, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.627500057220459, "explorer_return": 33.22500228881836, "explorer_subspace_norm": 8.386178016662598, "final_distance": 3.196431875228882, "max_param_abs": 0.36556291580200195, "policy_loss": -0.007972737543285646, "value_loss": 5.688779711723328}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:42.352779+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978162.352792, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16075065608779873, "elapsed_seconds": 578.7023619160755, "env_steps_per_second": 10368.024039394073, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9825914250828511, "eta_seconds": 7137.329130298264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0016772524867738614, "chaser_grad_norm": 1.004619687795639, "chaser_new_fields": 49.75, "chaser_partner_vision": 0.7115000486373901, "chaser_return": 15.110000610351562, "chaser_subspace_norm": 33.56522750854492, "collisions_per_episode": 14.199999809265137, "entropy": 0.43789007514715195, "explorer_grad_norm": 0.26522159203886986, "explorer_new_fields": 60.67499923706055, "explorer_partner_vision": 0.7115000486373901, "explorer_return": 30.387500762939453, "explorer_subspace_norm": 10.832931518554688, "final_distance": 3.0873048305511475, "max_param_abs": 0.4264068901538849, "policy_loss": -0.007891503338697703, "value_loss": 5.966992616653442}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:56.670440+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978356.6704519, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21377066938083672, "elapsed_seconds": 769.5744097710121, "env_steps_per_second": 10395.35605449824, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9239360244275303, "eta_seconds": 6926.169687939109, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0005306370660385085, "chaser_grad_norm": 0.2854561358690262, "chaser_new_fields": 45.025001525878906, "chaser_partner_vision": 0.749750018119812, "chaser_return": 10.557500839233398, "chaser_subspace_norm": 35.09560012817383, "collisions_per_episode": 10.925000190734863, "entropy": 0.37633907049894333, "explorer_grad_norm": 0.5534399151802063, "explorer_new_fields": 57.400001525878906, "explorer_partner_vision": 0.749750018119812, "explorer_return": 27.787500381469727, "explorer_subspace_norm": 14.880329132080078, "final_distance": 2.43146014213562, "max_param_abs": 0.49037474393844604, "policy_loss": -0.007005356951616193, "value_loss": 5.428126811981201}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:07.542489+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978547.542503, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2659455490066709, "elapsed_seconds": 957.4039764240151, "env_steps_per_second": 10444.911705245728, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.861618843046696, "eta_seconds": 6701.827834968106, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00042457543285778776, "chaser_grad_norm": 0.7817956060171127, "chaser_new_fields": 51.75, "chaser_partner_vision": 0.8770000338554382, "chaser_return": 15.405000686645508, "chaser_subspace_norm": 42.09107208251953, "collisions_per_episode": 14.300000190734863, "entropy": 0.2201327160000801, "explorer_grad_norm": 0.35337498039007187, "explorer_new_fields": 63.17499923706055, "explorer_partner_vision": 0.8770000338554382, "explorer_return": 33.41250228881836, "explorer_subspace_norm": 21.465106964111328, "final_distance": 1.9657284021377563, "max_param_abs": 0.48992887139320374, "policy_loss": -0.004997809281572785, "value_loss": 5.268904805183411}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:15.372057+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978735.3720722, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3181969677780742, "elapsed_seconds": 1145.509084001067, "env_steps_per_second": 10475.691696905671, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8031161507424205, "eta_seconds": 6491.218142672714, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005942620013130728, "chaser_grad_norm": 0.2406669482588768, "chaser_new_fields": 48.650001525878906, "chaser_partner_vision": 0.8715000152587891, "chaser_return": 14.960000991821289, "chaser_subspace_norm": 38.0869140625, "collisions_per_episode": 14.5, "entropy": 0.24184968322515488, "explorer_grad_norm": 0.33917396515607834, "explorer_new_fields": 62.67499923706055, "explorer_partner_vision": 0.8715000152587891, "explorer_return": 31.287500381469727, "explorer_subspace_norm": 23.78382682800293, "final_distance": 2.5971596240997314, "max_param_abs": 0.4951336979866028, "policy_loss": -0.006757004494780006, "value_loss": 5.6697787046432495}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:23.477164+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978923.477179, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3701415218894383, "elapsed_seconds": 1332.5094788019778, "env_steps_per_second": 10506.491865699154, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.744952888907352, "eta_seconds": 6281.8304000664675, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0011326391693771143, "chaser_grad_norm": 0.6869857050478458, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9332500696182251, "chaser_return": 18.27250099182129, "chaser_subspace_norm": 39.49469757080078, "collisions_per_episode": 16.875, "entropy": 0.20217590779066086, "explorer_grad_norm": 0.39582348614931107, "explorer_new_fields": 62.17499923706055, "explorer_partner_vision": 0.9332500696182251, "explorer_return": 29.98750114440918, "explorer_subspace_norm": 28.656511306762695, "final_distance": 2.0530452728271484, "max_param_abs": 0.5269516110420227, "policy_loss": -0.005966378787533699, "value_loss": 5.497568488121033}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:30.477560+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979110.4775732, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4221375967677967, "elapsed_seconds": 1519.6953483640682, "env_steps_per_second": 10528.42598829021, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6885503870711869, "eta_seconds": 6078.781393456273, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0008161310166217017, "chaser_grad_norm": 0.6782346665859222, "chaser_new_fields": 53.54999923706055, "chaser_partner_vision": 0.92125004529953, "chaser_return": 17.560001373291016, "chaser_subspace_norm": 39.778343200683594, "collisions_per_episode": 16.100000381469727, "entropy": 0.17546537518501282, "explorer_grad_norm": 0.31572260707616806, "explorer_new_fields": 61.125, "explorer_partner_vision": 0.92125004529953, "explorer_return": 28.274999618530273, "explorer_subspace_norm": 32.416229248046875, "final_distance": 2.000373601913452, "max_param_abs": 0.5484108924865723, "policy_loss": -0.005267302122433648, "value_loss": 5.505502820014954}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:37.663428+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979297.663442, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4748874584983504, "elapsed_seconds": 1709.5948505940614, "env_steps_per_second": 10528.810375010922, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6357234681609847, "eta_seconds": 5888.604485379545, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0005316504858008386, "chaser_grad_norm": 0.46923498064279556, "chaser_new_fields": 51.25, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 18.940000534057617, "chaser_subspace_norm": 36.8760986328125, "collisions_per_episode": 17.850000381469727, "entropy": 0.189016655087471, "explorer_grad_norm": 0.28533387184143066, "explorer_new_fields": 62.025001525878906, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 28.899999618530273, "explorer_subspace_norm": 32.292205810546875, "final_distance": 1.8484611511230469, "max_param_abs": 0.5896028876304626, "policy_loss": -0.005325202979147292, "value_loss": 5.30351197719574}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:04:47.562929+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979487.5629392, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5282632371133271, "elapsed_seconds": 1901.7476536079776, "env_steps_per_second": 10516.642395775376, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5847897113399814, "eta_seconds": 5705.242960823933, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0008929244335951214, "chaser_grad_norm": 0.635415144264698, "chaser_new_fields": 50.25, "chaser_partner_vision": 0.9347500205039978, "chaser_return": 19.135000228881836, "chaser_subspace_norm": 34.55233383178711, "collisions_per_episode": 18.100000381469727, "entropy": 0.2162095196545124, "explorer_grad_norm": 0.39251604676246643, "explorer_new_fields": 61.57500076293945, "explorer_partner_vision": 0.9347500205039978, "explorer_return": 26.712499618530273, "explorer_subspace_norm": 38.305789947509766, "final_distance": 2.1542279720306396, "max_param_abs": 0.5962355732917786, "policy_loss": -0.006229081626049293, "value_loss": 5.127236247062683}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:07:59.715734+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979679.7157483, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5802721394388937, "elapsed_seconds": 2088.979701980017, "env_steps_per_second": 10531.457045344929, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5298083676116287, "eta_seconds": 5507.310123401863, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0001547957173779757, "chaser_grad_norm": 0.45099158585071564, "chaser_new_fields": 44.54999923706055, "chaser_partner_vision": 0.9222500324249268, "chaser_return": 28.642499923706055, "chaser_subspace_norm": 38.59272384643555, "collisions_per_episode": 27.625, "entropy": 0.18400395289063454, "explorer_grad_norm": 0.4552285522222519, "explorer_new_fields": 55.85000228881836, "explorer_partner_vision": 0.9222500324249268, "explorer_return": 14.112500190734863, "explorer_subspace_norm": 39.83518981933594, "final_distance": 1.748358964920044, "max_param_abs": 0.6529774069786072, "policy_loss": -0.004401023029349638, "value_loss": 6.265259742736816}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:06.947780+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979866.9477918, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6331585321858357, "elapsed_seconds": 2279.3707158690086, "env_steps_per_second": 10529.221873788098, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4773699084336165, "eta_seconds": 5318.53167036102, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0009287971758098479, "chaser_grad_norm": 0.3374234735965729, "chaser_new_fields": 48.25, "chaser_partner_vision": 0.9342500567436218, "chaser_return": 18.61250114440918, "chaser_subspace_norm": 35.639488220214844, "collisions_per_episode": 17.92500114440918, "entropy": 0.21024496480822563, "explorer_grad_norm": 0.24169896915555, "explorer_new_fields": 61.04999923706055, "explorer_partner_vision": 0.9342500567436218, "explorer_return": 25.600000381469727, "explorer_subspace_norm": 40.536094665527344, "final_distance": 1.95449960231781, "max_param_abs": 0.7302749752998352, "policy_loss": -0.006967249414881138, "value_loss": 5.765379190444946}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:17.338796+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980057.3388112, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6871055653666716, "elapsed_seconds": 2473.5800353200175, "env_steps_per_second": 10511.080955032154, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4270654049923177, "eta_seconds": 5137.435457972344, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0005994347716713477, "chaser_grad_norm": 0.7940010726451874, "chaser_new_fields": 48.10000228881836, "chaser_partner_vision": 0.9465000629425049, "chaser_return": 18.962499618530273, "chaser_subspace_norm": 37.15826416015625, "collisions_per_episode": 18.225000381469727, "entropy": 0.20043616741895676, "explorer_grad_norm": 0.3976760804653168, "explorer_new_fields": 60.900001525878906, "explorer_partner_vision": 0.9465000629425049, "explorer_return": 25.67500114440918, "explorer_subspace_norm": 45.518123626708984, "final_distance": 1.8720436096191406, "max_param_abs": 0.7287337183952332, "policy_loss": -0.0057161247022428086, "value_loss": 5.595198750495911}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:31.548114+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980251.5481257, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7416190056802912, "elapsed_seconds": 2669.8284204490483, "env_steps_per_second": 10487.56533773454, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.377292439120541, "eta_seconds": 4958.252780833947, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.000874611195677083, "chaser_grad_norm": 0.40194160491228104, "chaser_new_fields": 47.125, "chaser_partner_vision": 0.937250018119812, "chaser_return": 19.010000228881836, "chaser_subspace_norm": 37.92008590698242, "collisions_per_episode": 18.5, "entropy": 0.19651446491479874, "explorer_grad_norm": 0.5923887342214584, "explorer_new_fields": 60.70000076293945, "explorer_partner_vision": 0.937250018119812, "explorer_return": 25.61250114440918, "explorer_subspace_norm": 52.10490036010742, "final_distance": 1.8134307861328125, "max_param_abs": 0.8093857169151306, "policy_loss": -0.00638780391775029, "value_loss": 5.844237327575684}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:20:47.796500+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980447.7965133, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7957362452716592, "elapsed_seconds": 2864.6504829779733, "env_steps_per_second": 10472.481783820702, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3262270754527654, "eta_seconds": 4774.417471629956, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0009340340659766966, "chaser_grad_norm": 0.5584994629025459, "chaser_new_fields": 49.650001525878906, "chaser_partner_vision": 0.9390000700950623, "chaser_return": 21.865001678466797, "chaser_subspace_norm": 38.4711799621582, "collisions_per_episode": 20.649999618530273, "entropy": 0.1979723945260048, "explorer_grad_norm": 0.49450917541980743, "explorer_new_fields": 61.47500228881836, "explorer_partner_vision": 0.9390000700950623, "explorer_return": 24.387500762939453, "explorer_subspace_norm": 54.4816780090332, "final_distance": 1.6306456327438354, "max_param_abs": 0.8194250464439392, "policy_loss": -0.005836165396495119, "value_loss": 5.313397645950317}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:02.618563+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980642.6185777, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8486546360614011, "elapsed_seconds": 3055.156689821044, "env_steps_per_second": 10474.094538789237, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2729819540921017, "eta_seconds": 4582.735034731566, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0006481464791119329, "chaser_grad_norm": 0.8932879418134689, "chaser_new_fields": 52.57500076293945, "chaser_partner_vision": 0.8952500224113464, "chaser_return": 20.285001754760742, "chaser_subspace_norm": 39.76984405517578, "collisions_per_episode": 18.75, "entropy": 0.1607821360230446, "explorer_grad_norm": 1.1095263063907623, "explorer_new_fields": 63.70000076293945, "explorer_partner_vision": 0.8952500224113464, "explorer_return": 29.725000381469727, "explorer_subspace_norm": 56.43265151977539, "final_distance": 1.9444500207901, "max_param_abs": 0.8587191700935364, "policy_loss": -0.004887927481906473, "value_loss": 5.260045766830444}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:13.124770+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980833.1247845, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9015540001125192, "elapsed_seconds": 3245.594400405069, "env_steps_per_second": 10475.739049758222, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2197495295639964, "eta_seconds": 4391.098306430387, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00041575367472290736, "chaser_grad_norm": 0.22038829699158669, "chaser_new_fields": 53.42499923706055, "chaser_partner_vision": 0.9005000591278076, "chaser_return": 18.817501068115234, "chaser_subspace_norm": 39.161991119384766, "collisions_per_episode": 17.325000762939453, "entropy": 0.20684056729078293, "explorer_grad_norm": 0.5319891422986984, "explorer_new_fields": 61.0, "explorer_partner_vision": 0.9005000591278076, "explorer_return": 26.875, "explorer_subspace_norm": 58.60102462768555, "final_distance": 1.9856764078140259, "max_param_abs": 0.8815340995788574, "policy_loss": -0.004912081582470584, "value_loss": 5.186847686767578}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:23.562479+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981023.5624905, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9554182859272179, "elapsed_seconds": 3439.5058293379843, "env_steps_per_second": 10466.619853622713, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1677334605777108, "eta_seconds": 4203.840458079759, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0011115005318802496, "chaser_grad_norm": 0.4741239733994007, "chaser_new_fields": 50.525001525878906, "chaser_partner_vision": 0.9430000185966492, "chaser_return": 25.095001220703125, "chaser_subspace_norm": 41.56173324584961, "collisions_per_episode": 23.55000114440918, "entropy": 0.11652559041976929, "explorer_grad_norm": 0.664473757147789, "explorer_new_fields": 65.32500457763672, "explorer_partner_vision": 0.9430000185966492, "explorer_return": 27.024999618530273, "explorer_subspace_norm": 60.8641471862793, "final_distance": 1.621573805809021, "max_param_abs": 0.8911052942276001, "policy_loss": -0.0036770399133683185, "value_loss": 4.6762083768844604}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:37.473909+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981217.473924, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.009048794256394, "elapsed_seconds": 3632.575659323018, "env_steps_per_second": 10460.897050409087, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1152644568096983, "eta_seconds": 4014.952044514914, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0004221649129567684, "chaser_grad_norm": 0.4742456451058388, "chaser_new_fields": 46.22500228881836, "chaser_partner_vision": 0.8830000162124634, "chaser_return": 18.942501068115234, "chaser_subspace_norm": 37.73898696899414, "collisions_per_episode": 18.524999618530273, "entropy": 0.18987196311354637, "explorer_grad_norm": 0.3590960279107094, "explorer_new_fields": 60.07500076293945, "explorer_partner_vision": 0.8830000162124634, "explorer_return": 23.649999618530273, "explorer_subspace_norm": 61.16921615600586, "final_distance": 1.8976906538009644, "max_param_abs": 0.902378261089325, "policy_loss": -0.004797064521349981, "value_loss": 5.6494282484054565}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:50.543737+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981410.5437493, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0615170873322253, "elapsed_seconds": 3821.461514396011, "env_steps_per_second": 10467.199486195028, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0615170873322253, "eta_seconds": 3821.461514396011, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0009190326497892087, "chaser_grad_norm": 0.5877722576260567, "chaser_new_fields": 45.650001525878906, "chaser_partner_vision": 0.9395000338554382, "chaser_return": 17.84250259399414, "chaser_subspace_norm": 37.249080657958984, "collisions_per_episode": 17.67500114440918, "entropy": 0.21706660464406013, "explorer_grad_norm": 0.5664015114307404, "explorer_new_fields": 59.82500076293945, "explorer_partner_vision": 0.9395000338554382, "explorer_return": 23.774999618530273, "explorer_subspace_norm": 70.35315704345703, "final_distance": 2.095780611038208, "max_param_abs": 0.9034669399261475, "policy_loss": -0.006788831242358739, "value_loss": 5.6421942710876465}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:39:59.429595+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981599.4296117, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.11473598729085, "elapsed_seconds": 4013.04955424706, "env_steps_per_second": 10465.856310084904, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.008570655167912, "eta_seconds": 3630.8543586044825, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005296420300595628, "chaser_grad_norm": 0.4503377303481102, "chaser_new_fields": 50.0, "chaser_partner_vision": 0.9190000295639038, "chaser_return": 17.657501220703125, "chaser_subspace_norm": 41.130897521972656, "collisions_per_episode": 16.625, "entropy": 0.18006031587719917, "explorer_grad_norm": 0.26743724197149277, "explorer_new_fields": 58.47500228881836, "explorer_partner_vision": 0.9190000295639038, "explorer_return": 23.587499618530273, "explorer_subspace_norm": 67.7103500366211, "final_distance": 1.8557966947555542, "max_param_abs": 0.9124955534934998, "policy_loss": -0.005980666249991273, "value_loss": 5.098747372627258}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:11.017633+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981791.0176437, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1692648428036287, "elapsed_seconds": 4209.353434093064, "env_steps_per_second": 10452.91175685753, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9566712350211508, "eta_seconds": 3444.016446076143, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0006415523072691443, "chaser_grad_norm": 0.6156454682350159, "chaser_new_fields": 48.375, "chaser_partner_vision": 0.9425000548362732, "chaser_return": 21.21000099182129, "chaser_subspace_norm": 42.44292068481445, "collisions_per_episode": 20.350000381469727, "entropy": 0.141556728631258, "explorer_grad_norm": 0.26799606531858444, "explorer_new_fields": 59.97500228881836, "explorer_partner_vision": 0.9425000548362732, "explorer_return": 23.30000114440918, "explorer_subspace_norm": 64.55814361572266, "final_distance": 1.6543800830841064, "max_param_abs": 0.924020528793335, "policy_loss": -0.0045606907354667214, "value_loss": 5.237063765525818}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:27.321513+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981987.3215277, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2227185409530648, "elapsed_seconds": 4401.786747431033, "env_steps_per_second": 10450.301806839343, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9037484867913956, "eta_seconds": 3253.4945524490245, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0008237254814575934, "chaser_grad_norm": 0.6340896785259247, "chaser_new_fields": 49.75, "chaser_partner_vision": 0.9352500438690186, "chaser_return": 16.825000762939453, "chaser_subspace_norm": 45.24385452270508, "collisions_per_episode": 15.949999809265137, "entropy": 0.14490226283669472, "explorer_grad_norm": 0.6062151193618774, "explorer_new_fields": 60.97500228881836, "explorer_partner_vision": 0.9352500438690186, "explorer_return": 27.149999618530273, "explorer_subspace_norm": 75.12525939941406, "final_distance": 1.674075961112976, "max_param_abs": 0.9924095273017883, "policy_loss": -0.0030751947443903305, "value_loss": 5.660175323486328}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:39.754825+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982179.754838, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2751079427263985, "elapsed_seconds": 4590.388593815034, "env_steps_per_second": 10456.63107142474, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8500719618175988, "eta_seconds": 3060.259062543356, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0003112658295007231, "chaser_grad_norm": 1.0334856808185577, "chaser_new_fields": 49.32500076293945, "chaser_partner_vision": 0.9395000338554382, "chaser_return": 20.86250114440918, "chaser_subspace_norm": 49.748783111572266, "collisions_per_episode": 19.774999618530273, "entropy": 0.1335170529782772, "explorer_grad_norm": 0.41471026837825775, "explorer_new_fields": 56.72500228881836, "explorer_partner_vision": 0.9395000338554382, "explorer_return": 19.575000762939453, "explorer_subspace_norm": 74.05712890625, "final_distance": 1.8723421096801758, "max_param_abs": 0.9675775766372681, "policy_loss": -0.0032862027799680504, "value_loss": 5.213485360145569}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:48.356674+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982368.356687, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.32713238714279, "elapsed_seconds": 4777.676593714044, "env_steps_per_second": 10465.337914622485, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7962794322856741, "eta_seconds": 2866.6059562284267, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005346971998765326, "chaser_grad_norm": 0.37228476256132126, "chaser_new_fields": 48.20000076293945, "chaser_partner_vision": 0.9350000619888306, "chaser_return": 20.045001983642578, "chaser_subspace_norm": 42.068748474121094, "collisions_per_episode": 19.350000381469727, "entropy": 0.1611567698419094, "explorer_grad_norm": 0.375806100666523, "explorer_new_fields": 58.375, "explorer_partner_vision": 0.9350000619888306, "explorer_return": 20.98750114440918, "explorer_subspace_norm": 105.0394287109375, "final_distance": 1.6606601476669312, "max_param_abs": 0.9974307417869568, "policy_loss": -0.004341715233401544, "value_loss": 5.334937810897827}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:55.644672+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982555.6446831, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3792829075252908, "elapsed_seconds": 4965.418467091047, "env_steps_per_second": 10472.430540273035, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.742690796359772, "eta_seconds": 2673.6868668951793, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00047676125171713934, "chaser_grad_norm": 0.3202666975557804, "chaser_new_fields": 44.5, "chaser_partner_vision": 0.9450000524520874, "chaser_return": 19.34000015258789, "chaser_subspace_norm": 40.434226989746094, "collisions_per_episode": 19.25, "entropy": 0.21454104408621788, "explorer_grad_norm": 0.4701062887907028, "explorer_new_fields": 56.54999923706055, "explorer_partner_vision": 0.9450000524520874, "explorer_return": 18.48750114440918, "explorer_subspace_norm": 86.6978988647461, "final_distance": 2.061654806137085, "max_param_abs": 0.9743969440460205, "policy_loss": -0.006256678724660958, "value_loss": 5.763511300086975}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:03.386547+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982743.3865619, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4314718463624982, "elapsed_seconds": 5153.298646904994, "env_steps_per_second": 10478.725123456938, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6892271852856473, "eta_seconds": 2481.21786702833, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0012956688417202766, "chaser_grad_norm": 0.3143876791000366, "chaser_new_fields": 49.04999923706055, "chaser_partner_vision": 0.9380000233650208, "chaser_return": 24.172500610351562, "chaser_subspace_norm": 46.76182174682617, "collisions_per_episode": 23.225000381469727, "entropy": 0.14750829711556435, "explorer_grad_norm": 0.37793220579624176, "explorer_new_fields": 58.17499923706055, "explorer_partner_vision": 0.9380000233650208, "explorer_return": 16.387500762939453, "explorer_subspace_norm": 93.05455017089844, "final_distance": 1.7932850122451782, "max_param_abs": 1.0117686986923218, "policy_loss": -0.005267503736547496, "value_loss": 5.2618759870529175}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:11.266725+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982931.2667375, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4850420674791611, "elapsed_seconds": 5346.15144292498, "env_steps_per_second": 10474.82485257869, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6364466003482119, "eta_seconds": 2291.2077612535627, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.001148998038539295, "chaser_grad_norm": 0.34996723011136055, "chaser_new_fields": 50.97500228881836, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 21.579999923706055, "chaser_subspace_norm": 44.12871170043945, "collisions_per_episode": 20.200000762939453, "entropy": 0.15114127099514008, "explorer_grad_norm": 0.41512512415647507, "explorer_new_fields": 59.20000076293945, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 20.899999618530273, "explorer_subspace_norm": 104.54307556152344, "final_distance": 1.9402074813842773, "max_param_abs": 1.0667240619659424, "policy_loss": -0.004499732237313481, "value_loss": 4.610663533210754}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:24.119522+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983124.1195357, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5381642108039038, "elapsed_seconds": 5537.391158894054, "env_steps_per_second": 10474.246506288704, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5834415972014808, "eta_seconds": 2100.3897499253308, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00033000036913377606, "chaser_grad_norm": 0.1884375885128975, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9025000333786011, "chaser_return": 17.1825008392334, "chaser_subspace_norm": 39.510009765625, "collisions_per_episode": 16.125, "entropy": 0.20632563903927803, "explorer_grad_norm": 0.2023741453886032, "explorer_new_fields": 63.47500228881836, "explorer_partner_vision": 0.9025000333786011, "explorer_return": 31.225000381469727, "explorer_subspace_norm": 97.64185333251953, "final_distance": 2.140202283859253, "max_param_abs": 1.0625193119049072, "policy_loss": -0.005696523889433536, "value_loss": 5.431702017784119}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:35.359238+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983315.3592489, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5920858503369446, "elapsed_seconds": 5731.509061213001, "env_steps_per_second": 10468.447202856165, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5306952834456482, "eta_seconds": 1910.5030204043335, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005167124319243044, "chaser_grad_norm": 0.9239912778139114, "chaser_new_fields": 49.17499923706055, "chaser_partner_vision": 0.9305000305175781, "chaser_return": 16.865001678466797, "chaser_subspace_norm": 46.21757888793945, "collisions_per_episode": 16.05000114440918, "entropy": 0.17099354416131973, "explorer_grad_norm": 0.32076873630285263, "explorer_new_fields": 57.79999923706055, "explorer_partner_vision": 0.9305000305175781, "explorer_return": 22.450000762939453, "explorer_subspace_norm": 96.14112854003906, "final_distance": 1.7582542896270752, "max_param_abs": 1.0959322452545166, "policy_loss": -0.004080008820167791, "value_loss": 5.437852144241333}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:49.477141+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983509.4771557, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.644112591734447, "elapsed_seconds": 5918.805330244009, "env_steps_per_second": 10475.086869843712, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4773230105035492, "eta_seconds": 1718.3628378127771, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005374289264183929, "chaser_grad_norm": 0.37107593566179276, "chaser_new_fields": 49.0, "chaser_partner_vision": 0.9472500681877136, "chaser_return": 20.7450008392334, "chaser_subspace_norm": 44.91938400268555, "collisions_per_episode": 20.05000114440918, "entropy": 0.16856927424669266, "explorer_grad_norm": 0.245488241314888, "explorer_new_fields": 59.150001525878906, "explorer_partner_vision": 0.9472500681877136, "explorer_return": 20.412500381469727, "explorer_subspace_norm": 85.51862335205078, "final_distance": 1.5606988668441772, "max_param_abs": 1.1283397674560547, "policy_loss": -0.0051685628555731, "value_loss": 5.328295826911926}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:14:56.773409+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983696.77342, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6988259566566881, "elapsed_seconds": 6115.773443964077, "env_steps_per_second": 10464.743435380913, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4247064891641721, "eta_seconds": 1528.9433609910195, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0010690009048202054, "chaser_grad_norm": 1.5101571381092072, "chaser_new_fields": 48.025001525878906, "chaser_partner_vision": 0.9257500171661377, "chaser_return": 12.982501029968262, "chaser_subspace_norm": 46.12118911743164, "collisions_per_episode": 12.625, "entropy": 0.1492394357919693, "explorer_grad_norm": 0.5180225670337677, "explorer_new_fields": 59.900001525878906, "explorer_partner_vision": 0.9257500171661377, "explorer_return": 29.412500381469727, "explorer_subspace_norm": 98.87821960449219, "final_distance": 1.7561384439468384, "max_param_abs": 1.1312024593353271, "policy_loss": -0.003736474832521708, "value_loss": 6.172814726829529}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:13.741523+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983893.741538, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7531581621138805, "elapsed_seconds": 6311.36938360997, "env_steps_per_second": 10457.318529223747, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3718820343877929, "eta_seconds": 1338.7753237960544, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 9.988718709674771e-05, "chaser_grad_norm": 0.17820751667022705, "chaser_new_fields": 43.25, "chaser_partner_vision": 0.9277500510215759, "chaser_return": 19.927501678466797, "chaser_subspace_norm": 45.344207763671875, "collisions_per_episode": 20.075000762939453, "entropy": 0.18734267354011536, "explorer_grad_norm": 0.4767478108406067, "explorer_new_fields": 53.10000228881836, "explorer_partner_vision": 0.9277500510215759, "explorer_return": 14.324999809265137, "explorer_subspace_norm": 100.08351135253906, "final_distance": 1.8032009601593018, "max_param_abs": 1.1664170026779175, "policy_loss": -0.004952042637391685, "value_loss": 5.541820168495178}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:29.337464+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984089.3374774, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8064167827655586, "elapsed_seconds": 6503.100417956011, "env_steps_per_second": 10456.550818782078, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31877943225274563, "eta_seconds": 1147.6059561098843, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0007968684427970274, "chaser_grad_norm": 0.6107524856925011, "chaser_new_fields": 48.375, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 21.68000030517578, "chaser_subspace_norm": 51.79354476928711, "collisions_per_episode": 20.80000114440918, "entropy": 0.1482456550002098, "explorer_grad_norm": 0.3292434960603714, "explorer_new_fields": 57.70000076293945, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 18.6875, "explorer_subspace_norm": 121.13251495361328, "final_distance": 1.563720703125, "max_param_abs": 1.2278400659561157, "policy_loss": -0.0041583591112868135, "value_loss": 5.81746518611908}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:41.068498+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984281.0685122, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8604707691880564, "elapsed_seconds": 6697.694769077003, "env_steps_per_second": 10451.35713308216, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2657815384554366, "eta_seconds": 956.8135384395719, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00041366020197797226, "chaser_grad_norm": 0.38722700253129005, "chaser_new_fields": 49.85000228881836, "chaser_partner_vision": 0.9197500348091125, "chaser_return": 16.617502212524414, "chaser_subspace_norm": 47.43393325805664, "collisions_per_episode": 15.77500057220459, "entropy": 0.16412756964564323, "explorer_grad_norm": 0.48008259385824203, "explorer_new_fields": 61.60000228881836, "explorer_partner_vision": 0.9197500348091125, "explorer_return": 28.475000381469727, "explorer_subspace_norm": 123.63157653808594, "final_distance": 1.892912745475769, "max_param_abs": 1.2527174949645996, "policy_loss": -0.004711385454985617, "value_loss": 5.507048606872559}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:55.662848+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984475.6628587, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9138005632872228, "elapsed_seconds": 6889.682027834002, "env_steps_per_second": 10450.409715444526, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21264450703191365, "eta_seconds": 765.5202253148891, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0008473068496036018, "chaser_grad_norm": 0.3696581646800041, "chaser_new_fields": 49.25, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 20.32000160217285, "chaser_subspace_norm": 45.366146087646484, "collisions_per_episode": 19.350000381469727, "entropy": 0.15908918529748917, "explorer_grad_norm": 0.4021201729774475, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 21.25, "explorer_subspace_norm": 145.34490966796875, "final_distance": 1.6058098077774048, "max_param_abs": 1.2972749471664429, "policy_loss": -0.0057415668405589315, "value_loss": 5.305572152137756}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:31:07.650109+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984667.6501255, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.967645665227222, "elapsed_seconds": 7083.524394818, "env_steps_per_second": 10446.777038579157, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15953883772112612, "eta_seconds": 574.339815796054, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0006182553142837222, "chaser_grad_norm": 0.4450978860259056, "chaser_new_fields": 48.17499923706055, "chaser_partner_vision": 0.9340000152587891, "chaser_return": 24.162500381469727, "chaser_subspace_norm": 49.698211669921875, "collisions_per_episode": 23.225000381469727, "entropy": 0.1645904928445816, "explorer_grad_norm": 0.32413702458143234, "explorer_new_fields": 58.375, "explorer_partner_vision": 0.9340000152587891, "explorer_return": 19.05000114440918, "explorer_subspace_norm": 120.44252014160156, "final_distance": 1.7177238464355469, "max_param_abs": 1.39788019657135, "policy_loss": -0.004847436100241964, "value_loss": 5.77062726020813}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:21.492475+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984861.492487, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0220887295052914, "elapsed_seconds": 7279.519426219049, "env_steps_per_second": 10440.249630527338, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10642572260554164, "eta_seconds": 383.1326013799499, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -0.0002089189239782574, "chaser_grad_norm": 0.31210391968488693, "chaser_new_fields": 52.57500076293945, "chaser_partner_vision": 0.9457500576972961, "chaser_return": 22.847501754760742, "chaser_subspace_norm": 54.08397674560547, "collisions_per_episode": 21.024999618530273, "entropy": 0.11483445204794407, "explorer_grad_norm": 0.4330414906144142, "explorer_new_fields": 63.35000228881836, "explorer_partner_vision": 0.9457500576972961, "explorer_return": 26.48750114440918, "explorer_subspace_norm": 111.18950653076172, "final_distance": 1.7784626483917236, "max_param_abs": 1.453328251838684, "policy_loss": -0.004174061445519683, "value_loss": 4.451533317565918}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:37.487506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985057.48752, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.073955157007796, "elapsed_seconds": 7466.2385652280645, "env_steps_per_second": 10447.027551900546, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05317833735917425, "eta_seconds": 191.4420144930273, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00018618514512080964, "chaser_grad_norm": 0.45402127504348755, "chaser_new_fields": 46.75, "chaser_partner_vision": 0.9292500615119934, "chaser_return": 19.457502365112305, "chaser_subspace_norm": 51.6260986328125, "collisions_per_episode": 18.975000381469727, "entropy": 0.14893564581871033, "explorer_grad_norm": 0.48403704911470413, "explorer_new_fields": 56.75, "explorer_partner_vision": 0.9292500615119934, "explorer_return": 19.11250114440918, "explorer_subspace_norm": 129.19778442382812, "final_distance": 1.6989506483078003, "max_param_abs": 1.4175773859024048, "policy_loss": -0.003678427097383974, "value_loss": 5.291145920753479}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:44.206643+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985244.206655, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.122798552864438, "elapsed_seconds": 7642.074790311977, "env_steps_per_second": 10468.361301752468, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:40.042868+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985420.0428798, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1228015855661297, "elapsed_seconds": 7642.0857080380665, "env_steps_per_second": 10468.3463463194, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.000801136315581541, "chaser_grad_norm": 0.38031933456659317, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 21.880001068115234, "chaser_subspace_norm": 52.41925811767578, "collisions_per_episode": 20.200000762939453, "entropy": 0.12773804366588593, "explorer_grad_norm": 0.23243309557437897, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.887500762939453, "explorer_subspace_norm": 119.99281311035156, "final_distance": 1.4895528554916382, "max_param_abs": 1.4659936428070068, "policy_loss": -0.0027591757117768623, "value_loss": 5.230631113052368}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:40.053787+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985420.0537992, "update": 20000} diff --git a/social/seed_0003/attempt_01/status.json b/social/seed_0003/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..47542ccc587783fdbb4cabd984dd088bf93be341 --- /dev/null +++ b/social/seed_0003/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1228015855661297, + "elapsed_seconds": 7642.0857080380665, + "env_steps_per_second": 10468.3463463194, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.000801136315581541, + "chaser_grad_norm": 0.38031933456659317, + "chaser_new_fields": 52.5, + "chaser_partner_vision": 0.9297500252723694, + "chaser_return": 21.880001068115234, + "chaser_subspace_norm": 52.41925811767578, + "collisions_per_episode": 20.200000762939453, + "entropy": 0.12773804366588593, + "explorer_grad_norm": 0.23243309557437897, + "explorer_new_fields": 59.29999923706055, + "explorer_partner_vision": 0.9297500252723694, + "explorer_return": 22.887500762939453, + "explorer_subspace_norm": 119.99281311035156, + "final_distance": 1.4895528554916382, + "max_param_abs": 1.4659936428070068, + "policy_loss": -0.0027591757117768623, + "value_loss": 5.230631113052368 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:40.053787+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985420.0537992, + "update": 20000 +} diff --git a/social/seed_0004/attempt_01/checkpoints/latest.safetensors b/social/seed_0004/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4d50fb8fc1dc13303d4b02c4ed6a57693cdc6e30 --- /dev/null +++ b/social/seed_0004/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5f51e56c6790b8377549c656b1f50cce6f1bc65fa30c4ac612097554d1d0555f +size 11460400 diff --git a/social/seed_0004/attempt_01/config.json b/social/seed_0004/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..db60783e72fba045de9941872360ebbb1d66122f --- /dev/null +++ b/social/seed_0004/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0004.attempt_01", + "seed": 4, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0004/attempt_01/metrics.jsonl b/social/seed_0004/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..c8cc4600aeb122dbab1fc99a041e81f4a04e0181 --- /dev/null +++ b/social/seed_0004/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005650202702640348, "elapsed_seconds": 2.0340729729505256, "env_steps_per_second": 1966.497787047334, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 11.299840385010434, "eta_seconds": 40679.42538603756, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -2.056581993722939e-05, "chaser_grad_norm": 0.10842846799641848, "chaser_new_fields": 36.72500228881836, "chaser_partner_vision": 0.3775000274181366, "chaser_return": -0.8250002264976501, "chaser_subspace_norm": 0.0022616556379944086, "collisions_per_episode": 1.7000000476837158, "entropy": 2.7724971175193787, "explorer_grad_norm": 0.10476464033126831, "explorer_new_fields": 34.625, "explorer_partner_vision": 0.3775000274181366, "explorer_return": 0.675000011920929, "explorer_subspace_norm": 0.0024901707656681538, "final_distance": 5.050706386566162, "max_param_abs": 0.1542823314666748, "policy_loss": -0.0008555040949724813, "value_loss": 3.204777419567108}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:36:20.015259+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977780.0152748, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05392308416443282, "elapsed_seconds": 194.12310299195815, "env_steps_per_second": 10302.740730879688, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.10300028241288, "eta_seconds": 7570.801016686368, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0008227154204858422, "chaser_grad_norm": 0.6484979093074799, "chaser_new_fields": 60.25, "chaser_partner_vision": 0.4595000147819519, "chaser_return": 8.83750057220459, "chaser_subspace_norm": 14.704169273376465, "collisions_per_episode": 6.375, "entropy": 0.6428284794092178, "explorer_grad_norm": 0.3108683004975319, "explorer_new_fields": 64.95000457763672, "explorer_partner_vision": 0.4595000147819519, "explorer_return": 42.36249923706055, "explorer_subspace_norm": 7.630646228790283, "final_distance": 4.210299015045166, "max_param_abs": 0.31605613231658936, "policy_loss": -0.005613177511840206, "value_loss": 5.577442407608032}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:39:32.104289+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782977972.1043, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10884240499112316, "elapsed_seconds": 391.83265796804335, "env_steps_per_second": 10208.439543408931, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0680056948313394, "eta_seconds": 7444.820501392823, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00048486648763845785, "chaser_grad_norm": 0.6231241077184677, "chaser_new_fields": 53.25, "chaser_partner_vision": 0.5815000534057617, "chaser_return": 14.092501640319824, "chaser_subspace_norm": 26.06329345703125, "collisions_per_episode": 12.52500057220459, "entropy": 0.54571932554245, "explorer_grad_norm": 0.2877832017838955, "explorer_new_fields": 62.875, "explorer_partner_vision": 0.5815000534057617, "explorer_return": 34.82500076293945, "explorer_subspace_norm": 8.353567123413086, "final_distance": 3.6115269660949707, "max_param_abs": 0.3698638379573822, "policy_loss": -0.006543429025917646, "value_loss": 6.0114076137542725}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:42:49.813843+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978169.8138576, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16109774514305172, "elapsed_seconds": 579.9518825149862, "env_steps_per_second": 10345.685876526071, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.986872190097638, "eta_seconds": 7152.739884351497, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00024907036182586495, "chaser_grad_norm": 0.46292533725500107, "chaser_new_fields": 49.04999923706055, "chaser_partner_vision": 0.7245000600814819, "chaser_return": 12.547500610351562, "chaser_subspace_norm": 32.998313903808594, "collisions_per_episode": 12.02500057220459, "entropy": 0.452880896627903, "explorer_grad_norm": 0.32380542904138565, "explorer_new_fields": 62.22500228881836, "explorer_partner_vision": 0.7245000600814819, "explorer_return": 33.23749923706055, "explorer_subspace_norm": 10.499018669128418, "final_distance": 3.1381642818450928, "max_param_abs": 0.4109547734260559, "policy_loss": -0.0067221808610469225, "value_loss": 5.7546316385269165}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:45:57.933066+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978357.9330788, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.213055311797232, "elapsed_seconds": 766.9991224700352, "env_steps_per_second": 10430.259651714974, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.917497806175088, "eta_seconds": 6902.992102230317, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0007705109087128292, "chaser_grad_norm": 0.3652050979435444, "chaser_new_fields": 48.72500228881836, "chaser_partner_vision": 0.8350000381469727, "chaser_return": 15.045001029968262, "chaser_subspace_norm": 36.56180953979492, "collisions_per_episode": 14.449999809265137, "entropy": 0.362516850233078, "explorer_grad_norm": 0.32036517560482025, "explorer_new_fields": 59.45000076293945, "explorer_partner_vision": 0.8350000381469727, "explorer_return": 28.3125, "explorer_subspace_norm": 13.671789169311523, "final_distance": 2.2175824642181396, "max_param_abs": 0.44999924302101135, "policy_loss": -0.006352211766875726, "value_loss": 5.786903500556946}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:49:04.980308+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978544.9803226, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26560944601498904, "elapsed_seconds": 956.1940056539606, "env_steps_per_second": 10458.128727925663, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8592661221049234, "eta_seconds": 6693.358039577724, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0008142175692160691, "chaser_grad_norm": 0.42791494354605675, "chaser_new_fields": 51.95000076293945, "chaser_partner_vision": 0.8345000147819519, "chaser_return": 14.665000915527344, "chaser_subspace_norm": 37.647850036621094, "collisions_per_episode": 13.5, "entropy": 0.2937430292367935, "explorer_grad_norm": 0.47259580343961716, "explorer_new_fields": 60.875, "explorer_partner_vision": 0.8345000147819519, "explorer_return": 30.700000762939453, "explorer_subspace_norm": 18.367706298828125, "final_distance": 2.559105634689331, "max_param_abs": 0.45478522777557373, "policy_loss": -0.006354553538374219, "value_loss": 5.781802415847778}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:52:14.175189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978734.1751997, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.318291058852499, "elapsed_seconds": 1145.8478118689964, "env_steps_per_second": 10472.594942976553, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8036493334974941, "eta_seconds": 6493.137600590979, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00027079761911513733, "chaser_grad_norm": 0.9446615725755692, "chaser_new_fields": 44.400001525878906, "chaser_partner_vision": 0.9142500162124634, "chaser_return": 21.842500686645508, "chaser_subspace_norm": 38.16933822631836, "collisions_per_episode": 21.475000381469727, "entropy": 0.26352914422750473, "explorer_grad_norm": 0.5925700664520264, "explorer_new_fields": 59.67499923706055, "explorer_partner_vision": 0.9142500162124634, "explorer_return": 23.5625, "explorer_subspace_norm": 25.10439682006836, "final_distance": 1.733415961265564, "max_param_abs": 0.4892232418060303, "policy_loss": -0.0033908244483171757, "value_loss": 6.184085249900818}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:55:23.828996+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782978923.8290107, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3724345813888875, "elapsed_seconds": 1340.7644929999951, "env_steps_per_second": 10441.803965642497, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7557630265476125, "eta_seconds": 6320.746895571405, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": -0.00012843816483076864, "chaser_grad_norm": 0.6736382991075516, "chaser_new_fields": 49.10000228881836, "chaser_partner_vision": 0.8800000548362732, "chaser_return": 16.700000762939453, "chaser_subspace_norm": 37.142616271972656, "collisions_per_episode": 16.05000114440918, "entropy": 0.2550524175167084, "explorer_grad_norm": 0.32974129170179367, "explorer_new_fields": 60.400001525878906, "explorer_partner_vision": 0.8800000548362732, "explorer_return": 27.25, "explorer_subspace_norm": 27.21210479736328, "final_distance": 1.9331449270248413, "max_param_abs": 0.4890628457069397, "policy_loss": -0.006383774680552179, "value_loss": 5.321722865104675}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T07:58:38.745679+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979118.745692, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4265587168563959, "elapsed_seconds": 1535.6113806830253, "env_steps_per_second": 10419.30282704948, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7062348674255836, "eta_seconds": 6142.445522732101, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0007779727199133113, "chaser_grad_norm": 0.40583793073892593, "chaser_new_fields": 48.47500228881836, "chaser_partner_vision": 0.8397500514984131, "chaser_return": 17.410001754760742, "chaser_subspace_norm": 41.3942756652832, "collisions_per_episode": 16.80000114440918, "entropy": 0.20803559944033623, "explorer_grad_norm": 0.4517665505409241, "explorer_new_fields": 58.04999923706055, "explorer_partner_vision": 0.8397500514984131, "explorer_return": 24.0625, "explorer_subspace_norm": 32.633060455322266, "final_distance": 1.9385753870010376, "max_param_abs": 0.575741708278656, "policy_loss": -0.005309362623378888, "value_loss": 5.410590410232544}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:01:53.592566+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979313.592579, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4784619328536145, "elapsed_seconds": 1722.4629582730122, "env_steps_per_second": 10450.152157726101, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6480355464957832, "eta_seconds": 5932.92796738482, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0006476162262016305, "chaser_grad_norm": 0.27134721353650093, "chaser_new_fields": 46.42499923706055, "chaser_partner_vision": 0.9175000190734863, "chaser_return": 18.22500228881836, "chaser_subspace_norm": 41.062801361083984, "collisions_per_episode": 17.80000114440918, "entropy": 0.23204779997467995, "explorer_grad_norm": 0.2324732355773449, "explorer_new_fields": 60.17499923706055, "explorer_partner_vision": 0.9175000190734863, "explorer_return": 25.0625, "explorer_subspace_norm": 32.80967330932617, "final_distance": 1.9229042530059814, "max_param_abs": 0.5940806865692139, "policy_loss": -0.0058804776468299735, "value_loss": 5.697959661483765}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:05:00.444143+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979500.4441547, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.531851483882767, "elapsed_seconds": 1914.6653419779614, "env_steps_per_second": 10445.689678248853, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5955544516483011, "eta_seconds": 5743.996025933884, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0011254392871866692, "chaser_grad_norm": 0.4692191556096077, "chaser_new_fields": 47.900001525878906, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 22.32750129699707, "chaser_subspace_norm": 47.57111358642578, "collisions_per_episode": 21.475000381469727, "entropy": 0.18224558979272842, "explorer_grad_norm": 0.44907645136117935, "explorer_new_fields": 58.150001525878906, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 19.73750114440918, "explorer_subspace_norm": 35.297908782958984, "final_distance": 1.6389917135238647, "max_param_abs": 0.6235131025314331, "policy_loss": -0.004895277009811849, "value_loss": 5.706926465034485}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:08:12.646527+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979692.6465416, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5862033643363975, "elapsed_seconds": 2110.332111611031, "env_steps_per_second": 10424.899416995158, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5454452332505024, "eta_seconds": 5563.602839701809, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005485947504668132, "chaser_grad_norm": 0.3692334294319153, "chaser_new_fields": 48.42499923706055, "chaser_partner_vision": 0.9360000491142273, "chaser_return": 18.96750259399414, "chaser_subspace_norm": 42.83165740966797, "collisions_per_episode": 18.274999618530273, "entropy": 0.20433569699525833, "explorer_grad_norm": 0.29782792925834656, "explorer_new_fields": 59.54999923706055, "explorer_partner_vision": 0.9360000491142273, "explorer_return": 24.337499618530273, "explorer_subspace_norm": 38.80984878540039, "final_distance": 1.7042158842086792, "max_param_abs": 0.6311439275741577, "policy_loss": -0.007648931517266533, "value_loss": 5.519146680831909}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:11:28.313298+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782979888.3133123, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6399261232705451, "elapsed_seconds": 2303.734043773962, "env_steps_per_second": 10417.869226207795, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4931609542979383, "eta_seconds": 5375.379435472578, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0008376683008499697, "chaser_grad_norm": 0.30171437934041023, "chaser_new_fields": 50.0, "chaser_partner_vision": 0.8940000534057617, "chaser_return": 18.2450008392334, "chaser_subspace_norm": 41.516571044921875, "collisions_per_episode": 17.05000114440918, "entropy": 0.18735594302415848, "explorer_grad_norm": 0.48069049417972565, "explorer_new_fields": 61.875, "explorer_partner_vision": 0.8940000534057617, "explorer_return": 29.600000381469727, "explorer_subspace_norm": 48.2872314453125, "final_distance": 1.8932923078536987, "max_param_abs": 0.6311439275741577, "policy_loss": -0.005792233524843482, "value_loss": 4.937587261199951}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:14:41.715229+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980081.715244, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.691766409617784, "elapsed_seconds": 2490.3590746240225, "env_steps_per_second": 10440.261512860472, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4367456199753974, "eta_seconds": 5172.284231911431, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00042947833976691996, "chaser_grad_norm": 0.2915813699364662, "chaser_new_fields": 48.32500076293945, "chaser_partner_vision": 0.9330000281333923, "chaser_return": 17.037500381469727, "chaser_subspace_norm": 39.13463592529297, "collisions_per_episode": 16.475000381469727, "entropy": 0.23400523886084557, "explorer_grad_norm": 0.5561796873807907, "explorer_new_fields": 60.45000076293945, "explorer_partner_vision": 0.9330000281333923, "explorer_return": 26.0625, "explorer_subspace_norm": 43.29941940307617, "final_distance": 1.6581863164901733, "max_param_abs": 0.6767386794090271, "policy_loss": -0.005664644482546777, "value_loss": 5.169563055038452}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:17:48.340258+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980268.3402696, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7439512352436173, "elapsed_seconds": 2678.2244468770223, "env_steps_per_second": 10454.687631818819, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3816237225952892, "eta_seconds": 4973.845401343041, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0011773707562988678, "chaser_grad_norm": 0.554933451116085, "chaser_new_fields": 49.60000228881836, "chaser_partner_vision": 0.9230000376701355, "chaser_return": 17.4950008392334, "chaser_subspace_norm": 40.59891891479492, "collisions_per_episode": 16.55000114440918, "entropy": 0.2045213095843792, "explorer_grad_norm": 0.19806089997291565, "explorer_new_fields": 61.275001525878906, "explorer_partner_vision": 0.9230000376701355, "explorer_return": 27.262500762939453, "explorer_subspace_norm": 54.04510498046875, "final_distance": 1.8361142873764038, "max_param_abs": 0.6974998116493225, "policy_loss": -0.006328542407608495, "value_loss": 5.630391001701355}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:20:56.205631+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980456.2056453, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7975608085355553, "elapsed_seconds": 2871.218910727999, "env_steps_per_second": 10448.524105183427, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3292680142259254, "eta_seconds": 4785.364851213331, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0011390541245308583, "chaser_grad_norm": 0.513223584741354, "chaser_new_fields": 48.10000228881836, "chaser_partner_vision": 0.9330000281333923, "chaser_return": 23.075002670288086, "chaser_subspace_norm": 40.80764389038086, "collisions_per_episode": 22.0, "entropy": 0.1820480152964592, "explorer_grad_norm": 0.2624660059809685, "explorer_new_fields": 59.0, "explorer_partner_vision": 0.9330000281333923, "explorer_return": 20.337499618530273, "explorer_subspace_norm": 59.452659606933594, "final_distance": 1.6288689374923706, "max_param_abs": 0.7281795740127563, "policy_loss": -0.005243872808293304, "value_loss": 5.722882866859436}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:24:09.200094+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980649.2001069, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8507387192505605, "elapsed_seconds": 3062.659389302018, "env_steps_per_second": 10448.435797913793, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2761080788758408, "eta_seconds": 4593.989083953027, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0003503556786093043, "chaser_grad_norm": 0.7736223116517067, "chaser_new_fields": 50.400001525878906, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 18.70250129699707, "chaser_subspace_norm": 46.74492263793945, "collisions_per_episode": 17.625, "entropy": 0.1525821089744568, "explorer_grad_norm": 0.27773766592144966, "explorer_new_fields": 63.10000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 28.412500381469727, "explorer_subspace_norm": 62.00983428955078, "final_distance": 1.4543617963790894, "max_param_abs": 0.7287328839302063, "policy_loss": -0.003618510363624594, "value_loss": 5.001829266548157}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:27:20.640575+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782980840.640589, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9043182612864055, "elapsed_seconds": 3255.5457406310597, "env_steps_per_second": 10443.71749278798, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2234894123286661, "eta_seconds": 4404.561884383198, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0016408601232873465, "chaser_grad_norm": 0.3725217692553997, "chaser_new_fields": 50.95000076293945, "chaser_partner_vision": 0.940500020980835, "chaser_return": 21.704999923706055, "chaser_subspace_norm": 45.66873550415039, "collisions_per_episode": 20.149999618530273, "entropy": 0.1446627676486969, "explorer_grad_norm": 0.2993212938308716, "explorer_new_fields": 60.97500228881836, "explorer_partner_vision": 0.940500020980835, "explorer_return": 24.11250114440918, "explorer_subspace_norm": 68.26232147216797, "final_distance": 1.7865352630615234, "max_param_abs": 0.8086429238319397, "policy_loss": -0.004827932525891754, "value_loss": 4.882992267608643}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:30:33.526927+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981033.52694, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9585392862366603, "elapsed_seconds": 3450.741430451977, "env_steps_per_second": 10432.540578760412, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1715480165114736, "eta_seconds": 4217.572859441305, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0013123999601063474, "chaser_grad_norm": 0.9313327372074127, "chaser_new_fields": 48.275001525878906, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 20.04250144958496, "chaser_subspace_norm": 45.23548126220703, "collisions_per_episode": 19.475000381469727, "entropy": 0.20291095599532127, "explorer_grad_norm": 0.39903828501701355, "explorer_new_fields": 61.45000076293945, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 23.962499618530273, "explorer_subspace_norm": 64.14064025878906, "final_distance": 2.0815083980560303, "max_param_abs": 0.7962862849235535, "policy_loss": -0.005230811689976633, "value_loss": 5.764302730560303}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:33:48.722616+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981228.722631, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0106772549791558, "elapsed_seconds": 3638.4381179249613, "env_steps_per_second": 10444.041857628677, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.117064334450646, "eta_seconds": 4021.4316040223252, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005643152491494607, "chaser_grad_norm": 0.3390261381864548, "chaser_new_fields": 53.0, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 20.442502975463867, "chaser_subspace_norm": 46.59177017211914, "collisions_per_episode": 18.625, "entropy": 0.14677583426237106, "explorer_grad_norm": 0.6298786401748657, "explorer_new_fields": 64.05000305175781, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 30.912500381469727, "explorer_subspace_norm": 61.59165573120117, "final_distance": 1.6749080419540405, "max_param_abs": 0.7865360975265503, "policy_loss": -0.0037148486301319394, "value_loss": 4.687524318695068}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:36:56.419302+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981416.4193134, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0628969342430679, "elapsed_seconds": 3826.428963275044, "env_steps_per_second": 10453.611025817649, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0628969342430679, "eta_seconds": 3826.4289632750447, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0002738917167799393, "chaser_grad_norm": 1.2391781955957413, "chaser_new_fields": 46.07500076293945, "chaser_partner_vision": 0.9257500171661377, "chaser_return": 14.205000877380371, "chaser_subspace_norm": 43.59848403930664, "collisions_per_episode": 14.100000381469727, "entropy": 0.20846595242619514, "explorer_grad_norm": 0.26415519788861275, "explorer_new_fields": 63.875, "explorer_partner_vision": 0.9257500171661377, "explorer_return": 34.22500228881836, "explorer_subspace_norm": 65.48961639404297, "final_distance": 2.013091564178467, "max_param_abs": 0.8095552921295166, "policy_loss": -0.003235120873432251, "value_loss": 5.754092454910278}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:40:04.410148+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981604.4101648, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1154436229830673, "elapsed_seconds": 4015.597042739042, "env_steps_per_second": 10459.216787188328, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.00921089698468, "eta_seconds": 3633.159229144848, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006714217457446381, "chaser_grad_norm": 0.50969098508358, "chaser_new_fields": 51.10000228881836, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 17.770002365112305, "chaser_subspace_norm": 46.356971740722656, "collisions_per_episode": 16.649999618530273, "entropy": 0.1692444011569023, "explorer_grad_norm": 0.6253417283296585, "explorer_new_fields": 59.20000076293945, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 24.399999618530273, "explorer_subspace_norm": 65.3006820678711, "final_distance": 1.6084197759628296, "max_param_abs": 0.7947602272033691, "policy_loss": -0.0048244935204291295, "value_loss": 5.53916323184967}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:43:13.578227+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981793.5782385, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1676217913255418, "elapsed_seconds": 4203.438448771951, "env_steps_per_second": 10467.620862357282, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9553269201754433, "eta_seconds": 3439.1769126315958, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0008927968385963325, "chaser_grad_norm": 0.25558965280652046, "chaser_new_fields": 50.04999923706055, "chaser_partner_vision": 0.9332500696182251, "chaser_return": 22.024999618530273, "chaser_subspace_norm": 52.54391860961914, "collisions_per_episode": 20.600000381469727, "entropy": 0.14776641875505447, "explorer_grad_norm": 0.9405873939394951, "explorer_new_fields": 61.625, "explorer_partner_vision": 0.9332500696182251, "explorer_return": 26.36250114440918, "explorer_subspace_norm": 78.59947204589844, "final_distance": 1.7253437042236328, "max_param_abs": 0.8336631655693054, "policy_loss": -0.003262170955537158, "value_loss": 5.188649535179138}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:46:21.419635+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782981981.4196513, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2215805913138966, "elapsed_seconds": 4397.690128730028, "env_steps_per_second": 10460.036667768576, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9029073935798366, "eta_seconds": 3250.466616887412, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0004046895241609505, "chaser_grad_norm": 0.24088818952441216, "chaser_new_fields": 50.29999923706055, "chaser_partner_vision": 0.9537500739097595, "chaser_return": 17.77250099182129, "chaser_subspace_norm": 46.08091735839844, "collisions_per_episode": 16.67500114440918, "entropy": 0.16840653494000435, "explorer_grad_norm": 0.532878689467907, "explorer_new_fields": 67.67500305175781, "explorer_partner_vision": 0.9537500739097595, "explorer_return": 35.29999923706055, "explorer_subspace_norm": 65.37627410888672, "final_distance": 1.714156150817871, "max_param_abs": 0.847284197807312, "policy_loss": -0.005025150935164646, "value_loss": 4.762189626693726}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:49:35.671314+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982175.6713266, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2752070368605524, "elapsed_seconds": 4590.7453326979885, "env_steps_per_second": 10455.818504702444, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8501380245737015, "eta_seconds": 3060.4968884653254, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0008162618304790037, "chaser_grad_norm": 0.5273931547999382, "chaser_new_fields": 47.79999923706055, "chaser_partner_vision": 0.9345000386238098, "chaser_return": 17.462501525878906, "chaser_subspace_norm": 49.142303466796875, "collisions_per_episode": 16.875, "entropy": 0.19980589300394058, "explorer_grad_norm": 0.8215000480413437, "explorer_new_fields": 62.20000076293945, "explorer_partner_vision": 0.9345000386238098, "explorer_return": 28.48750114440918, "explorer_subspace_norm": 75.5937271118164, "final_distance": 1.9811521768569946, "max_param_abs": 0.9113523364067078, "policy_loss": -0.0052181861856950285, "value_loss": 5.271326422691345}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:52:48.726517+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982368.7265313, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.327431592334437, "elapsed_seconds": 4778.753732403973, "env_steps_per_second": 10462.979010815709, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7964589554006622, "eta_seconds": 2867.252239442384, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00014226054980920333, "chaser_grad_norm": 0.48114994168281555, "chaser_new_fields": 48.900001525878906, "chaser_partner_vision": 0.9427500367164612, "chaser_return": 19.290000915527344, "chaser_subspace_norm": 57.75238037109375, "collisions_per_episode": 18.55000114440918, "entropy": 0.11706721223890781, "explorer_grad_norm": 0.24310096725821495, "explorer_new_fields": 60.67499923706055, "explorer_partner_vision": 0.9427500367164612, "explorer_return": 24.05000114440918, "explorer_subspace_norm": 73.70208740234375, "final_distance": 1.674404501914978, "max_param_abs": 0.9103676080703735, "policy_loss": -0.0026191213033168737, "value_loss": 5.154261112213135}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:55:56.734916+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982556.7349272, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3795217509375006, "elapsed_seconds": 4966.278303375002, "env_steps_per_second": 10470.617396665355, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7428194043509618, "eta_seconds": 2674.1498556634624, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 2.253567184018479e-05, "chaser_grad_norm": 0.5234654694795609, "chaser_new_fields": 52.125, "chaser_partner_vision": 0.9465000629425049, "chaser_return": 17.472501754760742, "chaser_subspace_norm": 58.50865936279297, "collisions_per_episode": 16.024999618530273, "entropy": 0.11500475741922855, "explorer_grad_norm": 0.5963674634695053, "explorer_new_fields": 64.05000305175781, "explorer_partner_vision": 0.9465000629425049, "explorer_return": 30.337499618530273, "explorer_subspace_norm": 82.48316955566406, "final_distance": 1.7131595611572266, "max_param_abs": 0.9014288783073425, "policy_loss": -0.0037908009327942693, "value_loss": 5.048547029495239}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T08:59:04.259489+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982744.2595038, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.432906848552488, "elapsed_seconds": 5158.464654788957, "env_steps_per_second": 10468.231075280915, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6899181122660129, "eta_seconds": 2483.705204157646, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0004510649674633882, "chaser_grad_norm": 0.29469920322299004, "chaser_new_fields": 52.85000228881836, "chaser_partner_vision": 0.9465000629425049, "chaser_return": 18.079999923706055, "chaser_subspace_norm": 55.534706115722656, "collisions_per_episode": 16.600000381469727, "entropy": 0.11153648421168327, "explorer_grad_norm": 0.529433012008667, "explorer_new_fields": 66.92500305175781, "explorer_partner_vision": 0.9465000629425049, "explorer_return": 33.98749923706055, "explorer_subspace_norm": 96.7128677368164, "final_distance": 1.7242405414581299, "max_param_abs": 0.9201875925064087, "policy_loss": -0.003193725266498859, "value_loss": 4.745298862457275}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:02:16.445840+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782982936.4458537, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.486663259596939, "elapsed_seconds": 5351.98773454898, "env_steps_per_second": 10463.402155894366, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6371413969701166, "eta_seconds": 2293.7090290924198, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0007471729956405326, "chaser_grad_norm": 0.5765634700655937, "chaser_new_fields": 50.07500076293945, "chaser_partner_vision": 0.9515000581741333, "chaser_return": 19.58500099182129, "chaser_subspace_norm": 51.26042938232422, "collisions_per_episode": 18.450000762939453, "entropy": 0.13874564692378044, "explorer_grad_norm": 0.6059492379426956, "explorer_new_fields": 60.92499923706055, "explorer_partner_vision": 0.9515000581741333, "explorer_return": 27.587499618530273, "explorer_subspace_norm": 105.52851104736328, "final_distance": 2.009634494781494, "max_param_abs": 0.9360404014587402, "policy_loss": -0.004843998290598339, "value_loss": 5.541916728019714}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:05:29.968920+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983129.9689355, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.540804535215009, "elapsed_seconds": 5546.896326774033, "env_steps_per_second": 10456.297825514195, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5844430995643138, "eta_seconds": 2103.99515843153, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.000307883327537381, "chaser_grad_norm": 0.47599275410175323, "chaser_new_fields": 50.025001525878906, "chaser_partner_vision": 0.9382500648498535, "chaser_return": 21.75750160217285, "chaser_subspace_norm": 54.17022705078125, "collisions_per_episode": 20.524999618530273, "entropy": 0.14921684190630913, "explorer_grad_norm": 0.806442379951477, "explorer_new_fields": 57.67499923706055, "explorer_partner_vision": 0.9382500648498535, "explorer_return": 19.6875, "explorer_subspace_norm": 94.66766357421875, "final_distance": 1.6523113250732422, "max_param_abs": 0.9466206431388855, "policy_loss": -0.003427367775236334, "value_loss": 5.719408631324768}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:08:44.877512+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983324.877525, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5943362496585987, "elapsed_seconds": 5739.610498770955, "env_steps_per_second": 10453.67103096073, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5314454165528663, "eta_seconds": 1913.2034995903184, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00035342174552632777, "chaser_grad_norm": 0.6859694197773933, "chaser_new_fields": 48.900001525878906, "chaser_partner_vision": 0.9505000710487366, "chaser_return": 19.135000228881836, "chaser_subspace_norm": 53.450565338134766, "collisions_per_episode": 18.100000381469727, "entropy": 0.13705725595355034, "explorer_grad_norm": 0.6459543257951736, "explorer_new_fields": 58.275001525878906, "explorer_partner_vision": 0.9505000710487366, "explorer_return": 23.225000381469727, "explorer_subspace_norm": 86.29521179199219, "final_distance": 1.7990089654922485, "max_param_abs": 0.9286988973617554, "policy_loss": -0.004262221205515004, "value_loss": 5.447039365768433}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:11:57.591684+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983517.5917, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6462593813472388, "elapsed_seconds": 5926.53377285006, "env_steps_per_second": 10461.426927832103, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47794627200403705, "eta_seconds": 1720.6065792145334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": -0.00023261655478901133, "chaser_grad_norm": 2.4417392015457153, "chaser_new_fields": 52.625, "chaser_partner_vision": 0.9437500238418579, "chaser_return": 10.162501335144043, "chaser_subspace_norm": 57.83631896972656, "collisions_per_episode": 9.175000190734863, "entropy": 0.1291857548058033, "explorer_grad_norm": 1.0154740735888481, "explorer_new_fields": 53.20000076293945, "explorer_partner_vision": 0.9437500238418579, "explorer_return": 22.0625, "explorer_subspace_norm": 99.2213134765625, "final_distance": 2.4684901237487793, "max_param_abs": 1.0182329416275024, "policy_loss": -0.002108115272129396, "value_loss": 6.3929678201675415}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:15:04.514958+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983704.5149732, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6982475842494427, "elapsed_seconds": 6113.691303297994, "env_steps_per_second": 10468.307414454437, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.42456189606236067, "eta_seconds": 1528.4228258244984, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0007729527072593001, "chaser_grad_norm": 0.28837253898382187, "chaser_new_fields": 44.22500228881836, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 26.837499618530273, "chaser_subspace_norm": 65.45987701416016, "collisions_per_episode": 25.975000381469727, "entropy": 0.18702024221420288, "explorer_grad_norm": 0.34226949512958527, "explorer_new_fields": 51.625, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 8.824999809265137, "explorer_subspace_norm": 82.82646179199219, "final_distance": 1.7187751531600952, "max_param_abs": 1.0881530046463013, "policy_loss": -0.004812586042568512, "value_loss": 6.719958782196045}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:18:11.672489+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782983891.672504, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7501309439161237, "elapsed_seconds": 6300.471398098045, "env_steps_per_second": 10475.406652893267, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3712398971943292, "eta_seconds": 1336.4636298995852, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0003544627506757436, "chaser_grad_norm": 0.4115583226084709, "chaser_new_fields": 45.775001525878906, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 20.66750144958496, "chaser_subspace_norm": 70.88333129882812, "collisions_per_episode": 20.17500114440918, "entropy": 0.18634789809584618, "explorer_grad_norm": 0.2553623095154762, "explorer_new_fields": 55.10000228881836, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 17.274999618530273, "explorer_subspace_norm": 77.9788818359375, "final_distance": 1.4543617963790894, "max_param_abs": 1.0996918678283691, "policy_loss": -0.004571893624028434, "value_loss": 5.433372855186462}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:21:18.452582+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984078.4525933, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8037721023988948, "elapsed_seconds": 6493.579568636022, "env_steps_per_second": 10471.8821539417, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3183127239527461, "eta_seconds": 1145.9258062298861, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0008394011314065697, "chaser_grad_norm": 0.5239869654178619, "chaser_new_fields": 48.5, "chaser_partner_vision": 0.9407500624656677, "chaser_return": 22.045001983642578, "chaser_subspace_norm": 60.25279235839844, "collisions_per_episode": 21.25, "entropy": 0.17551859468221664, "explorer_grad_norm": 0.27089823782444, "explorer_new_fields": 55.150001525878906, "explorer_partner_vision": 0.9407500624656677, "explorer_return": 17.037500381469727, "explorer_subspace_norm": 86.1933364868164, "final_distance": 1.9641268253326416, "max_param_abs": 1.0928677320480347, "policy_loss": -0.004979392812588301, "value_loss": 5.165536522865295}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:24:31.560754+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984271.5607686, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8578554509169447, "elapsed_seconds": 6688.279623301001, "env_steps_per_second": 10466.069593760718, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26540792155956355, "eta_seconds": 955.4685176144288, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0010362882793376604, "chaser_grad_norm": 0.3623509742319584, "chaser_new_fields": 47.375, "chaser_partner_vision": 0.9352500438690186, "chaser_return": 18.545001983642578, "chaser_subspace_norm": 58.46084976196289, "collisions_per_episode": 17.950000762939453, "entropy": 0.173718374222517, "explorer_grad_norm": 0.32294443994760513, "explorer_new_fields": 56.875, "explorer_partner_vision": 0.9352500438690186, "explorer_return": 21.012500762939453, "explorer_subspace_norm": 96.76380920410156, "final_distance": 1.5503047704696655, "max_param_abs": 1.090191125869751, "policy_loss": -0.004752456039188502, "value_loss": 5.306207895278931}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:27:46.260808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984466.2608213, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9120486712405627, "elapsed_seconds": 6883.375216466025, "env_steps_per_second": 10459.984780106948, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21244985236006253, "eta_seconds": 764.8194684962251, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 5.5630089089930834e-05, "chaser_grad_norm": 1.5480476021766663, "chaser_new_fields": 47.54999923706055, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 24.915002822875977, "chaser_subspace_norm": 60.93098449707031, "collisions_per_episode": 23.899999618530273, "entropy": 0.11976535804569721, "explorer_grad_norm": 0.4005112908780575, "explorer_new_fields": 53.75, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 11.625, "explorer_subspace_norm": 101.18220520019531, "final_distance": 1.6143203973770142, "max_param_abs": 1.112443447113037, "policy_loss": -0.002218019879428823, "value_loss": 4.89696991443634}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:31:01.356401+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984661.356415, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9659202835561074, "elapsed_seconds": 7077.313020801987, "env_steps_per_second": 10455.945608523398, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15939894190995466, "eta_seconds": 573.8361908758368, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005792167408837923, "chaser_grad_norm": 0.24662825465202332, "chaser_new_fields": 45.025001525878906, "chaser_partner_vision": 0.9510000348091125, "chaser_return": 19.047502517700195, "chaser_subspace_norm": 62.59393310546875, "collisions_per_episode": 18.725000381469727, "entropy": 0.14697032421827316, "explorer_grad_norm": 0.28777964413166046, "explorer_new_fields": 54.07500076293945, "explorer_partner_vision": 0.9510000348091125, "explorer_return": 15.71250057220459, "explorer_subspace_norm": 100.9203872680664, "final_distance": 1.5997438430786133, "max_param_abs": 1.111131191253662, "policy_loss": -0.004536015760153944, "value_loss": 5.1664347648620605}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:34:15.294206+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782984855.2942195, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.019953168310005, "elapsed_seconds": 7271.831405916018, "env_steps_per_second": 10451.287407209413, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.106313324647895, "eta_seconds": 382.727968732422, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0010951837727093405, "chaser_grad_norm": 0.4830399751663208, "chaser_new_fields": 47.900001525878906, "chaser_partner_vision": 0.8857500553131104, "chaser_return": 16.42750358581543, "chaser_subspace_norm": 54.02366638183594, "collisions_per_episode": 15.975000381469727, "entropy": 0.20296068862080574, "explorer_grad_norm": 0.21627317741513252, "explorer_new_fields": 54.125, "explorer_partner_vision": 0.8857500553131104, "explorer_return": 18.850000381469727, "explorer_subspace_norm": 94.59818267822266, "final_distance": 2.0229849815368652, "max_param_abs": 1.111131191253662, "policy_loss": -0.0054129646471707815, "value_loss": 5.0956220626831055}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:37:29.812591+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985049.8126056, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.0726598367797364, "elapsed_seconds": 7461.575412407052, "env_steps_per_second": 10453.556479547495, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05314512401999324, "eta_seconds": 191.32244647197567, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 4.0124208213761225e-05, "chaser_grad_norm": 0.3509816378355026, "chaser_new_fields": 47.54999923706055, "chaser_partner_vision": 0.9337500333786011, "chaser_return": 21.392499923706055, "chaser_subspace_norm": 67.07545471191406, "collisions_per_episode": 20.575000762939453, "entropy": 0.1040275264531374, "explorer_grad_norm": 0.5037448331713676, "explorer_new_fields": 62.875, "explorer_partner_vision": 0.9337500333786011, "explorer_return": 27.012500762939453, "explorer_subspace_norm": 100.067626953125, "final_distance": 1.5702221393585205, "max_param_abs": 1.097406029701233, "policy_loss": -0.002814068018923699, "value_loss": 4.53323233127594}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:40:39.556597+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985239.5566077, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1222522115761078, "elapsed_seconds": 7640.107961673988, "env_steps_per_second": 10471.056220843191, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:38.089145+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985418.0891554, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.122254997810556, "elapsed_seconds": 7640.117992118001, "env_steps_per_second": 10471.042473759273, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.000236154562574864, "chaser_grad_norm": 0.38722849637269974, "chaser_new_fields": 45.57500076293945, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 58.249393463134766, "collisions_per_episode": 19.125, "entropy": 0.19220001995563507, "explorer_grad_norm": 0.3825383335351944, "explorer_new_fields": 56.85000228881836, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 19.30000114440918, "explorer_subspace_norm": 90.54417419433594, "final_distance": 1.830224633216858, "max_param_abs": 1.0662928819656372, "policy_loss": -0.004796283046713867, "value_loss": 5.130034923553467}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T09:43:38.099176+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985418.0991888, "update": 20000} diff --git a/social/seed_0004/attempt_01/status.json b/social/seed_0004/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..2950f95c68b950a5660b338ee03643ed78d6fc91 --- /dev/null +++ b/social/seed_0004/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.122254997810556, + "elapsed_seconds": 7640.117992118001, + "env_steps_per_second": 10471.042473759273, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.000236154562574864, + "chaser_grad_norm": 0.38722849637269974, + "chaser_new_fields": 45.57500076293945, + "chaser_partner_vision": 0.9225000143051147, + "chaser_return": 19.41750144958496, + "chaser_subspace_norm": 58.249393463134766, + "collisions_per_episode": 19.125, + "entropy": 0.19220001995563507, + "explorer_grad_norm": 0.3825383335351944, + "explorer_new_fields": 56.85000228881836, + "explorer_partner_vision": 0.9225000143051147, + "explorer_return": 19.30000114440918, + "explorer_subspace_norm": 90.54417419433594, + "final_distance": 1.830224633216858, + "max_param_abs": 1.0662928819656372, + "policy_loss": -0.004796283046713867, + "value_loss": 5.130034923553467 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T09:43:38.099176+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782985418.0991888, + "update": 20000 +} diff --git a/social/seed_0005/attempt_01/checkpoints/latest.safetensors b/social/seed_0005/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..e2ecbaa98b86e844543d3d6de101e260decd5b8f --- /dev/null +++ b/social/seed_0005/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3405d3fc0a24dd17d23126ea788a9e1e3877444eb16623f74e0636865ace9e9b +size 11460408 diff --git a/social/seed_0005/attempt_01/config.json b/social/seed_0005/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..36b83d5e151ce7b2541b51901ee80c8140b007b7 --- /dev/null +++ b/social/seed_0005/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0005.attempt_01", + "seed": 5, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0005/attempt_01/metrics.jsonl b/social/seed_0005/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..d22aeecf694cf142e3c5325b8fe480a5244b27ee --- /dev/null +++ b/social/seed_0005/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.00043777609363638073, "elapsed_seconds": 1.5759939370909706, "env_steps_per_second": 2538.080830046435, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 8.75508409663398, "eta_seconds": 31518.302747882324, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -3.827462441040406e-05, "chaser_grad_norm": 0.23730506375432014, "chaser_new_fields": 33.525001525878906, "chaser_partner_vision": 0.30525001883506775, "chaser_return": -1.9150002002716064, "chaser_subspace_norm": 0.002422536024823785, "collisions_per_episode": 1.3000000715255737, "entropy": 2.772470533847809, "explorer_grad_norm": 0.08085168339312077, "explorer_new_fields": 34.97500228881836, "explorer_partner_vision": 0.30525001883506775, "explorer_return": 1.5125000476837158, "explorer_subspace_norm": 0.0030809545423835516, "final_distance": 5.475633144378662, "max_param_abs": 0.15428578853607178, "policy_loss": -0.0009154173561860546, "value_loss": 3.1339523792266846}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:20.834300+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985400.8343134, "update": 1} +{"device": "cuda", "elapsed_hours": 0.04160537842528558, "elapsed_seconds": 149.7793623310281, "env_steps_per_second": 13352.97446105953, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6226097585861377, "eta_seconds": 5841.395130910096, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00032929822639915507, "chaser_grad_norm": 0.44155047088861465, "chaser_new_fields": 51.42499923706055, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 14.072501182556152, "chaser_subspace_norm": 15.31710433959961, "collisions_per_episode": 13.02500057220459, "entropy": 0.5538445711135864, "explorer_grad_norm": 0.23347657546401024, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 31.125, "explorer_subspace_norm": 9.003698348999023, "final_distance": 1.8160871267318726, "max_param_abs": 0.3036521375179291, "policy_loss": -0.003363415987230667, "value_loss": 5.539378523826599}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:45:49.037667+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985549.037679, "update": 500} +{"device": "cuda", "elapsed_hours": 0.09527903253196403, "elapsed_seconds": 343.0045171150705, "env_steps_per_second": 11661.654002819118, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8103016181073164, "eta_seconds": 6517.085825186339, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0002685723771342907, "chaser_grad_norm": 0.5023379810154438, "chaser_new_fields": 49.22500228881836, "chaser_partner_vision": 0.921500027179718, "chaser_return": 16.127500534057617, "chaser_subspace_norm": 18.77627182006836, "collisions_per_episode": 15.375, "entropy": 0.4405377432703972, "explorer_grad_norm": 0.539054349064827, "explorer_new_fields": 61.42499923706055, "explorer_partner_vision": 0.921500027179718, "explorer_return": 29.087499618530273, "explorer_subspace_norm": 13.400713920593262, "final_distance": 1.6054985523223877, "max_param_abs": 0.3279077410697937, "policy_loss": -0.005872102522282319, "value_loss": 5.701381325721741}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:49:02.262822+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985742.2628367, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.14802918346333577, "elapsed_seconds": 532.9050604680087, "env_steps_per_second": 11259.04114089416, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8256932627144744, "eta_seconds": 6572.4957457721075, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00017653192458588762, "chaser_grad_norm": 0.6680483222007751, "chaser_new_fields": 49.20000076293945, "chaser_partner_vision": 0.9122500419616699, "chaser_return": 13.992502212524414, "chaser_subspace_norm": 25.239545822143555, "collisions_per_episode": 13.52500057220459, "entropy": 0.3507041856646538, "explorer_grad_norm": 0.3137746602296829, "explorer_new_fields": 62.57500076293945, "explorer_partner_vision": 0.9122500419616699, "explorer_return": 32.337501525878906, "explorer_subspace_norm": 17.833385467529297, "final_distance": 1.8642282485961914, "max_param_abs": 0.3807339668273926, "policy_loss": -0.0054605520088237824, "value_loss": 5.412071704864502}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:52:12.163365+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985932.1633751, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.2011535510491942, "elapsed_seconds": 724.1527837770991, "env_steps_per_second": 11047.392455322624, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8103819594427477, "eta_seconds": 6517.375053993892, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002516953592905935, "chaser_grad_norm": 0.3246125802397728, "chaser_new_fields": 51.67499923706055, "chaser_partner_vision": 0.9337500333786011, "chaser_return": 17.065000534057617, "chaser_subspace_norm": 29.052703857421875, "collisions_per_episode": 15.949999809265137, "entropy": 0.2820078581571579, "explorer_grad_norm": 0.43635717034339905, "explorer_new_fields": 64.6500015258789, "explorer_partner_vision": 0.9337500333786011, "explorer_return": 32.625, "explorer_subspace_norm": 24.66944122314453, "final_distance": 1.6383394002914429, "max_param_abs": 0.40484559535980225, "policy_loss": -0.005391026715741987, "value_loss": 5.396326780319214}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:55:23.411089+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986123.4111028, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2543692006383531, "elapsed_seconds": 915.7291222980712, "env_steps_per_second": 10920.259885264395, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7805844044684718, "eta_seconds": 6410.103856086498, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00036908014438652437, "chaser_grad_norm": 0.3821464851498604, "chaser_new_fields": 50.900001525878906, "chaser_partner_vision": 0.8615000247955322, "chaser_return": 16.837499618530273, "chaser_subspace_norm": 29.196014404296875, "collisions_per_episode": 15.824999809265137, "entropy": 0.2635941579937935, "explorer_grad_norm": 0.3420739099383354, "explorer_new_fields": 62.45000076293945, "explorer_partner_vision": 0.8615000247955322, "explorer_return": 30.4375, "explorer_subspace_norm": 26.760404586791992, "final_distance": 2.379201889038086, "max_param_abs": 0.4318005442619324, "policy_loss": -0.005460368725010767, "value_loss": 5.691980957984924}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:58:34.987426+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986314.9874372, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3072574038347294, "elapsed_seconds": 1106.1266538050259, "env_steps_per_second": 10848.667246847855, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7411252883967998, "eta_seconds": 6268.051038228479, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.000893155890465791, "chaser_grad_norm": 0.5485154911875725, "chaser_new_fields": 50.275001525878906, "chaser_partner_vision": 0.8982500433921814, "chaser_return": 16.062501907348633, "chaser_subspace_norm": 32.38425827026367, "collisions_per_episode": 15.375, "entropy": 0.21320565044879913, "explorer_grad_norm": 0.6284236684441566, "explorer_new_fields": 65.1500015258789, "explorer_partner_vision": 0.8982500433921814, "explorer_return": 32.61249923706055, "explorer_subspace_norm": 30.584808349609375, "final_distance": 2.1404948234558105, "max_param_abs": 0.43318766355514526, "policy_loss": -0.005004348743707254, "value_loss": 5.626531958580017}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:01:45.384958+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986505.3849714, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.36036110740336075, "elapsed_seconds": 1297.2999866520986, "env_steps_per_second": 10791.644295109692, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6988452206158435, "eta_seconds": 6115.8427942170365, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 7.209374585492867e-05, "chaser_grad_norm": 0.28207409381866455, "chaser_new_fields": 50.625, "chaser_partner_vision": 0.8937500715255737, "chaser_return": 14.52500057220459, "chaser_subspace_norm": 31.404247283935547, "collisions_per_episode": 13.75, "entropy": 0.22709926962852478, "explorer_grad_norm": 0.5671822875738144, "explorer_new_fields": 64.25, "explorer_partner_vision": 0.8937500715255737, "explorer_return": 34.17499923706055, "explorer_subspace_norm": 34.387081146240234, "final_distance": 2.0039238929748535, "max_param_abs": 0.49748802185058594, "policy_loss": -0.005624466077893686, "value_loss": 5.3764872550964355}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:04:56.558291+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986696.5583014, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.41410585122638927, "elapsed_seconds": 1490.7810644150013, "env_steps_per_second": 10732.628943257052, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6564234049055573, "eta_seconds": 5963.124257660006, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0010362966226686865, "chaser_grad_norm": 0.45977988839149475, "chaser_new_fields": 49.57500076293945, "chaser_partner_vision": 0.9265000224113464, "chaser_return": 17.212499618530273, "chaser_subspace_norm": 34.71195983886719, "collisions_per_episode": 16.375, "entropy": 0.1997036710381508, "explorer_grad_norm": 0.49618013948202133, "explorer_new_fields": 62.72500228881836, "explorer_partner_vision": 0.9265000224113464, "explorer_return": 29.11250114440918, "explorer_subspace_norm": 38.30999755859375, "final_distance": 1.9345086812973022, "max_param_abs": 0.5606271624565125, "policy_loss": -0.005907846817745366, "value_loss": 5.345959782600403}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:08:10.039369+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986890.0393834, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.46838637055612503, "elapsed_seconds": 1686.1909340020502, "env_steps_per_second": 10674.947680615458, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.613330831915542, "eta_seconds": 5807.990994895951, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00033735177582183695, "chaser_grad_norm": 0.6195323765277863, "chaser_new_fields": 47.60000228881836, "chaser_partner_vision": 0.9357500672340393, "chaser_return": 17.440000534057617, "chaser_subspace_norm": 33.69694900512695, "collisions_per_episode": 16.950000762939453, "entropy": 0.20496303215622902, "explorer_grad_norm": 0.3352973610162735, "explorer_new_fields": 63.525001525878906, "explorer_partner_vision": 0.9357500672340393, "explorer_return": 30.8125, "explorer_subspace_norm": 39.3932991027832, "final_distance": 1.8103021383285522, "max_param_abs": 0.6538880467414856, "policy_loss": -0.005816969985140474, "value_loss": 5.5162999629974365}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:11:25.449238+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987085.4492486, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5204012504105534, "elapsed_seconds": 1873.4445014779922, "env_steps_per_second": 10675.523072192243, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.56120375123166, "eta_seconds": 5620.3335044339765, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0009200741595081219, "chaser_grad_norm": 0.27518555521965027, "chaser_new_fields": 50.57500076293945, "chaser_partner_vision": 0.9507500529289246, "chaser_return": 19.41499900817871, "chaser_subspace_norm": 37.02577590942383, "collisions_per_episode": 18.200000762939453, "entropy": 0.15045235306024551, "explorer_grad_norm": 0.3475729823112488, "explorer_new_fields": 63.72500228881836, "explorer_partner_vision": 0.9507500529289246, "explorer_return": 29.92500114440918, "explorer_subspace_norm": 49.22797775268555, "final_distance": 1.9390543699264526, "max_param_abs": 0.7204680442810059, "policy_loss": -0.0034155924115335523, "value_loss": 5.560041546821594}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:14:32.702806+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987272.7028193, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.573339878596954, "elapsed_seconds": 2064.0235629490344, "env_steps_per_second": 10658.79304622223, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5115324072101515, "eta_seconds": 5441.516665956546, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005687297050913642, "chaser_grad_norm": 0.8723443448543549, "chaser_new_fields": 46.54999923706055, "chaser_partner_vision": 0.924250066280365, "chaser_return": 21.357501983642578, "chaser_subspace_norm": 36.10696792602539, "collisions_per_episode": 20.875, "entropy": 0.14181451871991158, "explorer_grad_norm": 0.37872714549303055, "explorer_new_fields": 62.17499923706055, "explorer_partner_vision": 0.924250066280365, "explorer_return": 25.73750114440918, "explorer_subspace_norm": 52.50918197631836, "final_distance": 1.7008377313613892, "max_param_abs": 0.7182598114013672, "policy_loss": -0.0040701838145054126, "value_loss": 5.573855519294739}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:17:43.281867+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987463.2818782, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6277473910316864, "elapsed_seconds": 2259.890607714071, "env_steps_per_second": 10619.983072665862, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4647439124072683, "eta_seconds": 5273.0780846661655, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0002784032543328152, "chaser_grad_norm": 0.37566038966178894, "chaser_new_fields": 48.75, "chaser_partner_vision": 0.9335000514984131, "chaser_return": 18.212501525878906, "chaser_subspace_norm": 35.46366500854492, "collisions_per_episode": 17.475000381469727, "entropy": 0.14615919813513756, "explorer_grad_norm": 1.5339600294828415, "explorer_new_fields": 62.47500228881836, "explorer_partner_vision": 0.9335000514984131, "explorer_return": 29.087499618530273, "explorer_subspace_norm": 54.796409606933594, "final_distance": 1.7509593963623047, "max_param_abs": 0.7178844809532166, "policy_loss": -0.00426072374917652, "value_loss": 5.393409013748169}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:20:59.148912+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987659.1489255, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6795418673627945, "elapsed_seconds": 2446.3507225060603, "env_steps_per_second": 10628.075427126574, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4113561860611887, "eta_seconds": 5080.882269820279, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00010179102275653662, "chaser_grad_norm": 0.4401996023952961, "chaser_new_fields": 50.60000228881836, "chaser_partner_vision": 0.9327500462532043, "chaser_return": 19.274999618530273, "chaser_subspace_norm": 35.656131744384766, "collisions_per_episode": 18.25, "entropy": 0.18866830691695213, "explorer_grad_norm": 0.37249017506837845, "explorer_new_fields": 62.32500076293945, "explorer_partner_vision": 0.9327500462532043, "explorer_return": 27.537500381469727, "explorer_subspace_norm": 55.25505828857422, "final_distance": 1.5313720703125, "max_param_abs": 0.810156524181366, "policy_loss": -0.0039042244013387517, "value_loss": 5.6979875564575195}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:24:05.609026+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987845.6090372, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.731261331267241, "elapsed_seconds": 2632.5407925620675, "env_steps_per_second": 10636.1124884031, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3580567580677332, "eta_seconds": 4889.00432904384, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00042420504139784043, "chaser_grad_norm": 1.4187753200531006, "chaser_new_fields": 47.35000228881836, "chaser_partner_vision": 0.9587500691413879, "chaser_return": 24.21500015258789, "chaser_subspace_norm": 32.7892951965332, "collisions_per_episode": 23.200000762939453, "entropy": 0.17944413051009178, "explorer_grad_norm": 0.3887479491531849, "explorer_new_fields": 56.0, "explorer_partner_vision": 0.9587500691413879, "explorer_return": 15.237500190734863, "explorer_subspace_norm": 52.22835922241211, "final_distance": 1.472299337387085, "max_param_abs": 0.8463520407676697, "policy_loss": -0.002151237820276819, "value_loss": 5.524796366691589}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:27:11.799097+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988031.799112, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7844215568527902, "elapsed_seconds": 2823.917604670045, "env_steps_per_second": 10623.539422817292, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3073692614213173, "eta_seconds": 4706.529341116742, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0008886806998880115, "chaser_grad_norm": 0.4323786571621895, "chaser_new_fields": 49.04999923706055, "chaser_partner_vision": 0.9447500705718994, "chaser_return": 18.10500144958496, "chaser_subspace_norm": 35.15358352661133, "collisions_per_episode": 17.200000762939453, "entropy": 0.19827055931091309, "explorer_grad_norm": 0.3535676673054695, "explorer_new_fields": 60.67499923706055, "explorer_partner_vision": 0.9447500705718994, "explorer_return": 26.11250114440918, "explorer_subspace_norm": 55.654296875, "final_distance": 1.735202670097351, "max_param_abs": 0.8134477734565735, "policy_loss": -0.005609304415994387, "value_loss": 5.427311301231384}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:30:23.175908+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988223.175919, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8370350568608329, "elapsed_seconds": 3013.3262046989985, "env_steps_per_second": 10619.494149056618, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2555525852912495, "eta_seconds": 4519.989307048498, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005373235877308424, "chaser_grad_norm": 0.42715349048376083, "chaser_new_fields": 50.17499923706055, "chaser_partner_vision": 0.9362500309944153, "chaser_return": 19.700000762939453, "chaser_subspace_norm": 38.8222541809082, "collisions_per_episode": 18.75, "entropy": 0.12904895842075348, "explorer_grad_norm": 0.37726913020014763, "explorer_new_fields": 63.54999923706055, "explorer_partner_vision": 0.9362500309944153, "explorer_return": 28.11250114440918, "explorer_subspace_norm": 72.3985595703125, "final_distance": 1.720725655555725, "max_param_abs": 0.8113468289375305, "policy_loss": -0.0038188930056062276, "value_loss": 5.18973970413208}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:33:32.584510+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988412.5845234, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8893106783800047, "elapsed_seconds": 3201.518442168017, "env_steps_per_second": 10619.96068870862, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2031850354553004, "eta_seconds": 4331.466127639082, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": -0.0003750752296570181, "chaser_grad_norm": 0.7244221791625023, "chaser_new_fields": 53.17499923706055, "chaser_partner_vision": 0.9547500610351562, "chaser_return": 23.252500534057617, "chaser_subspace_norm": 39.856903076171875, "collisions_per_episode": 21.67500114440918, "entropy": 0.09981817752122879, "explorer_grad_norm": 0.3658481277525425, "explorer_new_fields": 64.07500457763672, "explorer_partner_vision": 0.9547500610351562, "explorer_return": 25.1875, "explorer_subspace_norm": 81.31741333007812, "final_distance": 1.5025851726531982, "max_param_abs": 0.828362762928009, "policy_loss": -0.002350221858360868, "value_loss": 5.458203434944153}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:36:40.776746+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988600.7767565, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.941130098092512, "elapsed_seconds": 3388.0683531330433, "env_steps_per_second": 10625.523527797122, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1502701198908478, "eta_seconds": 4140.972431607052, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00030862065988390697, "chaser_grad_norm": 0.6697158291935921, "chaser_new_fields": 51.35000228881836, "chaser_partner_vision": 0.9325000643730164, "chaser_return": 25.315000534057617, "chaser_subspace_norm": 36.59380340576172, "collisions_per_episode": 23.80000114440918, "entropy": 0.1400493085384369, "explorer_grad_norm": 0.7113858759403229, "explorer_new_fields": 62.07500076293945, "explorer_partner_vision": 0.9325000643730164, "explorer_return": 21.837499618530273, "explorer_subspace_norm": 72.5894546508789, "final_distance": 1.6585254669189453, "max_param_abs": 0.9618297219276428, "policy_loss": -0.0032245467375027204, "value_loss": 5.289536476135254}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:39:47.326658+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988787.3266718, "update": 9000} +{"device": "cuda", "elapsed_hours": 0.9941844429541703, "elapsed_seconds": 3579.063994635013, "env_steps_per_second": 10617.301075633652, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.098835436949346, "eta_seconds": 3955.807573017646, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005237200430876499, "chaser_grad_norm": 0.5484417378902435, "chaser_new_fields": 50.75, "chaser_partner_vision": 0.9487500190734863, "chaser_return": 23.66499900817871, "chaser_subspace_norm": 37.69060516357422, "collisions_per_episode": 22.350000381469727, "entropy": 0.13963981717824936, "explorer_grad_norm": 0.31709080189466476, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.9487500190734863, "explorer_return": 21.25, "explorer_subspace_norm": 72.56075286865234, "final_distance": 1.5678894519805908, "max_param_abs": 0.9861094355583191, "policy_loss": -0.00419517484493781, "value_loss": 5.180540084838867}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:42:58.322301+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988978.3223147, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.047868602182522, "elapsed_seconds": 3772.3269678570796, "env_steps_per_second": 10603.534725602678, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.047868602182522, "eta_seconds": 3772.3269678570796, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006614649092253622, "chaser_grad_norm": 0.546450212597847, "chaser_new_fields": 48.875, "chaser_partner_vision": 0.9377500414848328, "chaser_return": 21.332502365112305, "chaser_subspace_norm": 40.512123107910156, "collisions_per_episode": 20.274999618530273, "entropy": 0.17647982761263847, "explorer_grad_norm": 0.2895813062787056, "explorer_new_fields": 58.35000228881836, "explorer_partner_vision": 0.9377500414848328, "explorer_return": 20.11250114440918, "explorer_subspace_norm": 80.26006317138672, "final_distance": 1.55388605594635, "max_param_abs": 0.9508216381072998, "policy_loss": -0.005359409181409092, "value_loss": 5.532748103141785}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:46:11.585274+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989171.5852897, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.101128493254461, "elapsed_seconds": 3964.0625757160597, "env_steps_per_second": 10595.190968299286, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9962591129445123, "eta_seconds": 3586.5328066002444, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0015871248698937634, "chaser_grad_norm": 0.41763633117079735, "chaser_new_fields": 47.650001525878906, "chaser_partner_vision": 0.9200000166893005, "chaser_return": 22.04250144958496, "chaser_subspace_norm": 40.13907241821289, "collisions_per_episode": 21.225000381469727, "entropy": 0.1493425816297531, "explorer_grad_norm": 0.3818689063191414, "explorer_new_fields": 60.75, "explorer_partner_vision": 0.9200000166893005, "explorer_return": 23.274999618530273, "explorer_subspace_norm": 84.37419128417969, "final_distance": 1.8582853078842163, "max_param_abs": 0.9477961659431458, "policy_loss": -0.005458723680114019, "value_loss": 5.429543495178223}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:49:23.320882+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989363.3208952, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1537161958430726, "elapsed_seconds": 4153.378305035061, "env_steps_per_second": 10593.785773537566, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9439496147806958, "eta_seconds": 3398.218613210505, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0012830375786494486, "chaser_grad_norm": 0.38080117106437683, "chaser_new_fields": 46.85000228881836, "chaser_partner_vision": 0.9352500438690186, "chaser_return": 23.517501831054688, "chaser_subspace_norm": 38.561702728271484, "collisions_per_episode": 22.67500114440918, "entropy": 0.21242719143629074, "explorer_grad_norm": 0.6976934671401978, "explorer_new_fields": 57.17499923706055, "explorer_partner_vision": 0.9352500438690186, "explorer_return": 18.649999618530273, "explorer_subspace_norm": 95.0623550415039, "final_distance": 1.6583225727081299, "max_param_abs": 0.9645247459411621, "policy_loss": -0.006433694863691386, "value_loss": 5.6233497858047485}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:52:32.636612+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989552.6366272, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2071908029202798, "elapsed_seconds": 4345.8868905130075, "env_steps_per_second": 10584.720946239344, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8922714630280328, "eta_seconds": 3212.1772669009183, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0014261145552767474, "chaser_grad_norm": 0.31601520627737045, "chaser_new_fields": 48.875, "chaser_partner_vision": 0.9437500238418579, "chaser_return": 23.915002822875977, "chaser_subspace_norm": 39.40815734863281, "collisions_per_episode": 22.75, "entropy": 0.15400954708456993, "explorer_grad_norm": 0.2398057021200657, "explorer_new_fields": 61.375, "explorer_partner_vision": 0.9437500238418579, "explorer_return": 21.087499618530273, "explorer_subspace_norm": 85.59361267089844, "final_distance": 1.7730686664581299, "max_param_abs": 1.0561943054199219, "policy_loss": -0.004380037242546209, "value_loss": 5.128291964530945}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:55:45.145196+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989745.1452115, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.259716476247219, "elapsed_seconds": 4534.979314489989, "env_steps_per_second": 10584.392269802041, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8398109841648127, "eta_seconds": 3023.3195429933257, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0010504067364318637, "chaser_grad_norm": 0.4034429341554642, "chaser_new_fields": 43.35000228881836, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 20.450000762939453, "chaser_subspace_norm": 41.10194396972656, "collisions_per_episode": 20.450000762939453, "entropy": 0.17071116715669632, "explorer_grad_norm": 0.42389291524887085, "explorer_new_fields": 59.625, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 21.412500381469727, "explorer_subspace_norm": 93.8042984008789, "final_distance": 1.7035683393478394, "max_param_abs": 1.0765948295593262, "policy_loss": -0.004693537459242947, "value_loss": 5.1131978034973145}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:58:54.237620+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989934.2376337, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3118865784569709, "elapsed_seconds": 4722.791682445095, "env_steps_per_second": 10586.958596089058, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7871319470741824, "eta_seconds": 2833.675009467057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0011002822802659296, "chaser_grad_norm": 0.5163806602358818, "chaser_new_fields": 50.82500076293945, "chaser_partner_vision": 0.9377500414848328, "chaser_return": 23.280000686645508, "chaser_subspace_norm": 43.95779800415039, "collisions_per_episode": 21.899999618530273, "entropy": 0.15090007334947586, "explorer_grad_norm": 0.3131614699959755, "explorer_new_fields": 63.47500228881836, "explorer_partner_vision": 0.9377500414848328, "explorer_return": 24.587499618530273, "explorer_subspace_norm": 95.73839569091797, "final_distance": 1.38800847530365, "max_param_abs": 1.1000360250473022, "policy_loss": -0.0037293717269797355, "value_loss": 5.261346459388733}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:02:02.049988+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990122.0500004, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.363699396931948, "elapsed_seconds": 4909.317828955012, "env_steps_per_second": 10592.102978810117, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7342996752710489, "eta_seconds": 2643.478830975776, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0010288599191383696, "chaser_grad_norm": 0.46028465405106544, "chaser_new_fields": 46.45000076293945, "chaser_partner_vision": 0.9307500720024109, "chaser_return": 21.032499313354492, "chaser_subspace_norm": 42.24143981933594, "collisions_per_episode": 20.42500114440918, "entropy": 0.15491418540477753, "explorer_grad_norm": 0.8786889463663101, "explorer_new_fields": 58.625, "explorer_partner_vision": 0.9307500720024109, "explorer_return": 19.774999618530273, "explorer_subspace_norm": 99.85923767089844, "final_distance": 1.5661664009094238, "max_param_abs": 1.0919440984725952, "policy_loss": -0.003918804021551203, "value_loss": 5.44774329662323}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:05:08.576134+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990308.5761468, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.416715213196973, "elapsed_seconds": 5100.174767509103, "env_steps_per_second": 10587.872467431798, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6821221396874315, "eta_seconds": 2455.6397028747533, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0004419732943656829, "chaser_grad_norm": 0.3612554110586643, "chaser_new_fields": 53.20000076293945, "chaser_partner_vision": 0.9542500376701355, "chaser_return": 20.702499389648438, "chaser_subspace_norm": 40.82073974609375, "collisions_per_episode": 19.024999618530273, "entropy": 0.14328867197036743, "explorer_grad_norm": 1.1671051532030106, "explorer_new_fields": 65.92500305175781, "explorer_partner_vision": 0.9542500376701355, "explorer_return": 30.412500381469727, "explorer_subspace_norm": 106.20968627929688, "final_distance": 1.3587192296981812, "max_param_abs": 1.0996874570846558, "policy_loss": -0.003738626637495912, "value_loss": 5.110748052597046}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:08:19.433071+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990499.4330816, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4697331609555597, "elapsed_seconds": 5291.039379440015, "env_steps_per_second": 10583.931810752623, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6298856404095255, "eta_seconds": 2267.588305474292, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0009241837841207357, "chaser_grad_norm": 0.702658161520958, "chaser_new_fields": 46.54999923706055, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 22.975000381469727, "chaser_subspace_norm": 37.68461608886719, "collisions_per_episode": 22.399999618530273, "entropy": 0.2082212008535862, "explorer_grad_norm": 0.44074392318725586, "explorer_new_fields": 58.025001525878906, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 17.0625, "explorer_subspace_norm": 98.20284271240234, "final_distance": 1.596573829650879, "max_param_abs": 1.0890860557556152, "policy_loss": -0.00627802232047614, "value_loss": 5.584610939025879}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:11:30.297684+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990690.2976975, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.522105085633359, "elapsed_seconds": 5479.578308280092, "env_steps_per_second": 10584.756113870524, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.577350204895412, "eta_seconds": 2078.460737623483, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.000513673315553298, "chaser_grad_norm": 0.5576435104012489, "chaser_new_fields": 48.625, "chaser_partner_vision": 0.940000057220459, "chaser_return": 20.252500534057617, "chaser_subspace_norm": 41.58379364013672, "collisions_per_episode": 19.42500114440918, "entropy": 0.15124722570180893, "explorer_grad_norm": 0.8982564210891724, "explorer_new_fields": 61.82500076293945, "explorer_partner_vision": 0.940000057220459, "explorer_return": 25.075000762939453, "explorer_subspace_norm": 113.86859130859375, "final_distance": 1.6337740421295166, "max_param_abs": 1.05161452293396, "policy_loss": -0.0031488226586962043, "value_loss": 5.607636570930481}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:14:38.836612+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990878.8366237, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.574745599874457, "elapsed_seconds": 5669.084159548045, "env_steps_per_second": 10583.720105644607, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5249151999581523, "eta_seconds": 1889.6947198493483, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00045423461125437825, "chaser_grad_norm": 0.3297133520245552, "chaser_new_fields": 46.32500076293945, "chaser_partner_vision": 0.9257500171661377, "chaser_return": 19.227500915527344, "chaser_subspace_norm": 40.37676239013672, "collisions_per_episode": 18.875, "entropy": 0.1905398815870285, "explorer_grad_norm": 0.49392949789762497, "explorer_new_fields": 59.25, "explorer_partner_vision": 0.9257500171661377, "explorer_return": 22.762500762939453, "explorer_subspace_norm": 105.80331420898438, "final_distance": 1.9580806493759155, "max_param_abs": 1.0589962005615234, "policy_loss": -0.004418823103419456, "value_loss": 5.757702946662903}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:17:48.342465+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991068.3424788, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6271708463461254, "elapsed_seconds": 5857.815046846052, "env_steps_per_second": 10584.151173120747, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4724044392617784, "eta_seconds": 1700.6559813424021, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0010107573849360563, "chaser_grad_norm": 0.24096596613526344, "chaser_new_fields": 47.775001525878906, "chaser_partner_vision": 0.9342500567436218, "chaser_return": 23.755002975463867, "chaser_subspace_norm": 39.879207611083984, "collisions_per_episode": 22.75, "entropy": 0.20127328857779503, "explorer_grad_norm": 0.3548465743660927, "explorer_new_fields": 59.17499923706055, "explorer_partner_vision": 0.9342500567436218, "explorer_return": 22.17500114440918, "explorer_subspace_norm": 111.14995574951172, "final_distance": 1.7752102613449097, "max_param_abs": 1.104758381843567, "policy_loss": -0.005178330790272412, "value_loss": 5.994641900062561}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:20:57.073350+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991257.073362, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6789375768361303, "elapsed_seconds": 6044.175276610069, "env_steps_per_second": 10588.706824514027, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4197343942090326, "eta_seconds": 1511.0438191525172, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005826382229474625, "chaser_grad_norm": 0.5693052522838116, "chaser_new_fields": 50.875, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 22.145002365112305, "chaser_subspace_norm": 46.48992156982422, "collisions_per_episode": 20.80000114440918, "entropy": 0.12821241095662117, "explorer_grad_norm": 0.3289678692817688, "explorer_new_fields": 61.650001525878906, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 23.825000762939453, "explorer_subspace_norm": 115.33250427246094, "final_distance": 1.6539961099624634, "max_param_abs": 1.122863531112671, "policy_loss": -0.003480862608647328, "value_loss": 5.587159037590027}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:24:03.433582+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991443.4335954, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.731432115573908, "elapsed_seconds": 6233.155616066069, "env_steps_per_second": 10588.537181693944, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.36727347906113206, "eta_seconds": 1322.1845246200753, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0008875713551277364, "chaser_grad_norm": 0.4042047709226608, "chaser_new_fields": 48.82500076293945, "chaser_partner_vision": 0.9517500400543213, "chaser_return": 22.20250129699707, "chaser_subspace_norm": 44.09364318847656, "collisions_per_episode": 21.125, "entropy": 0.14768925309181213, "explorer_grad_norm": 0.5328620299696922, "explorer_new_fields": 61.275001525878906, "explorer_partner_vision": 0.9517500400543213, "explorer_return": 24.1875, "explorer_subspace_norm": 104.33907318115234, "final_distance": 1.5753437280654907, "max_param_abs": 1.1366033554077148, "policy_loss": -0.003958008965943449, "value_loss": 5.014141082763672}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:27:12.413920+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991632.413931, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7831229676075033, "elapsed_seconds": 6419.242683387012, "env_steps_per_second": 10593.149901620618, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3146687589895594, "eta_seconds": 1132.807532362414, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00035164691924771674, "chaser_grad_norm": 0.292745690792799, "chaser_new_fields": 54.97500228881836, "chaser_partner_vision": 0.9512500166893005, "chaser_return": 21.962499618530273, "chaser_subspace_norm": 43.968631744384766, "collisions_per_episode": 19.774999618530273, "entropy": 0.11281907372176647, "explorer_grad_norm": 0.5271111354231834, "explorer_new_fields": 66.82500457763672, "explorer_partner_vision": 0.9512500166893005, "explorer_return": 31.3125, "explorer_subspace_norm": 115.09920501708984, "final_distance": 1.6249080896377563, "max_param_abs": 1.1564093828201294, "policy_loss": -0.0026133537702293097, "value_loss": 4.685871720314026}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:30:18.500989+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991818.5010026, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8357853083491662, "elapsed_seconds": 6608.827110056998, "env_steps_per_second": 10591.89457286261, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26225504404988087, "eta_seconds": 944.1181585795712, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00020651600975263307, "chaser_grad_norm": 0.606506809592247, "chaser_new_fields": 49.82500076293945, "chaser_partner_vision": 0.9497500658035278, "chaser_return": 19.740001678466797, "chaser_subspace_norm": 42.58631896972656, "collisions_per_episode": 18.55000114440918, "entropy": 0.1798299327492714, "explorer_grad_norm": 0.22327490150928497, "explorer_new_fields": 61.0, "explorer_partner_vision": 0.9497500658035278, "explorer_return": 24.80000114440918, "explorer_subspace_norm": 95.23554229736328, "final_distance": 1.7458511590957642, "max_param_abs": 1.253448247909546, "policy_loss": -0.004742766230350526, "value_loss": 4.8869709968566895}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:33:28.085414+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992008.0854254, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.8876700703391947, "elapsed_seconds": 6795.612253221101, "env_steps_per_second": 10595.071837106687, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.20974111892657718, "eta_seconds": 755.0680281356779, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00033098676616738487, "chaser_grad_norm": 0.33920105546712875, "chaser_new_fields": 50.5, "chaser_partner_vision": 0.9590000510215759, "chaser_return": 20.700000762939453, "chaser_subspace_norm": 47.248958587646484, "collisions_per_episode": 19.399999618530273, "entropy": 0.1348160095512867, "explorer_grad_norm": 0.35399428755044937, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.9590000510215759, "explorer_return": 22.912500381469727, "explorer_subspace_norm": 122.89472198486328, "final_distance": 1.649840235710144, "max_param_abs": 1.2641557455062866, "policy_loss": -0.002883355684694777, "value_loss": 5.135150194168091}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:36:34.870558+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992194.8705726, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9401223511569616, "elapsed_seconds": 6984.440464165062, "env_steps_per_second": 10594.978993617373, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15730721766137526, "eta_seconds": 566.305983580951, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.000932598889520353, "chaser_grad_norm": 0.23596013709902763, "chaser_new_fields": 48.60000228881836, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 23.465002059936523, "chaser_subspace_norm": 44.41228485107422, "collisions_per_episode": 22.450000762939453, "entropy": 0.10757111012935638, "explorer_grad_norm": 0.9262560606002808, "explorer_new_fields": 62.775001525878906, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 24.762500762939453, "explorer_subspace_norm": 144.63406372070312, "final_distance": 1.4690064191818237, "max_param_abs": 1.291568398475647, "policy_loss": -0.0021577100688610074, "value_loss": 5.429338455200195}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:39:43.698768+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992383.698779, "update": 18500} +{"device": "cuda", "elapsed_hours": 1.9936323461127985, "elapsed_seconds": 7177.076446006075, "env_steps_per_second": 10589.269958562689, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10492801821646308, "eta_seconds": 377.7408655792671, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00017163656335128508, "chaser_grad_norm": 0.4783521965146065, "chaser_new_fields": 50.650001525878906, "chaser_partner_vision": 0.9357500672340393, "chaser_return": 21.27250099182129, "chaser_subspace_norm": 46.32050323486328, "collisions_per_episode": 20.024999618530273, "entropy": 0.15996642410755157, "explorer_grad_norm": 0.306297205388546, "explorer_new_fields": 58.525001525878906, "explorer_partner_vision": 0.9357500672340393, "explorer_return": 23.012500762939453, "explorer_subspace_norm": 121.7342758178711, "final_distance": 1.6541683673858643, "max_param_abs": 1.3142790794372559, "policy_loss": -0.00367128911241954, "value_loss": 5.676693558692932}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:42:56.334753+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992576.3347676, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.04617606747695, "elapsed_seconds": 7366.2338429170195, "env_steps_per_second": 10588.857435608112, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05246605301222949, "eta_seconds": 188.87779084402615, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.000621079146096104, "chaser_grad_norm": 0.34488896280527115, "chaser_new_fields": 50.150001525878906, "chaser_partner_vision": 0.9220000505447388, "chaser_return": 17.505001068115234, "chaser_subspace_norm": 45.416053771972656, "collisions_per_episode": 16.450000762939453, "entropy": 0.19137008488178253, "explorer_grad_norm": 0.38153036683797836, "explorer_new_fields": 62.42499923706055, "explorer_partner_vision": 0.9220000505447388, "explorer_return": 29.375, "explorer_subspace_norm": 104.75499725341797, "final_distance": 1.8972313404083252, "max_param_abs": 1.3173565864562988, "policy_loss": -0.004619453563357467, "value_loss": 4.853449463844299}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:46:05.492146+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992765.4921575, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.099071822472777, "elapsed_seconds": 7556.658560901997, "env_steps_per_second": 10586.68978560424, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:49:15.916865+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992955.916876, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.0991321679530666, "elapsed_seconds": 7556.8758046310395, "env_steps_per_second": 10586.385441318756, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0005827252787078407, "chaser_grad_norm": 0.3728112056851387, "chaser_new_fields": 51.47500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 20.117502212524414, "chaser_subspace_norm": 50.84516143798828, "collisions_per_episode": 18.575000762939453, "entropy": 0.13566195219755173, "explorer_grad_norm": 0.6062372475862503, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.95250004529953, "explorer_return": 20.774999618530273, "explorer_subspace_norm": 104.22787475585938, "final_distance": 1.5808099508285522, "max_param_abs": 1.3590010404586792, "policy_loss": -0.003127762281341262, "value_loss": 4.4600828886032104}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:49:16.134110+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992956.134124, "update": 20000} diff --git a/social/seed_0005/attempt_01/status.json b/social/seed_0005/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..8184fa9784cd09f056f6a2d95e6ec8795b77033d --- /dev/null +++ b/social/seed_0005/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.0991321679530666, + "elapsed_seconds": 7556.8758046310395, + "env_steps_per_second": 10586.385441318756, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0005827252787078407, + "chaser_grad_norm": 0.3728112056851387, + "chaser_new_fields": 51.47500228881836, + "chaser_partner_vision": 0.95250004529953, + "chaser_return": 20.117502212524414, + "chaser_subspace_norm": 50.84516143798828, + "collisions_per_episode": 18.575000762939453, + "entropy": 0.13566195219755173, + "explorer_grad_norm": 0.6062372475862503, + "explorer_new_fields": 58.57500076293945, + "explorer_partner_vision": 0.95250004529953, + "explorer_return": 20.774999618530273, + "explorer_subspace_norm": 104.22787475585938, + "final_distance": 1.5808099508285522, + "max_param_abs": 1.3590010404586792, + "policy_loss": -0.003127762281341262, + "value_loss": 4.4600828886032104 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:49:16.134110+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782992956.134124, + "update": 20000 +} diff --git a/social/seed_0006/attempt_01/checkpoints/latest.safetensors b/social/seed_0006/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..144dd55d08f8a9a4aaf178b2d2c02b6e1d0718ae --- /dev/null +++ b/social/seed_0006/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6a45c928ba31f02d9b235a4605d844df712be9103fbf93bfe039029aa9400fc4 +size 11460416 diff --git a/social/seed_0006/attempt_01/config.json b/social/seed_0006/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..e2721e9e4bb8381b988ab3b5f3f425675feadd4a --- /dev/null +++ b/social/seed_0006/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0006.attempt_01", + "seed": 6, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0006/attempt_01/metrics.jsonl b/social/seed_0006/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..c4fb7e5dd09671d2db6f789a379d5f57a30b8c98 --- /dev/null +++ b/social/seed_0006/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.000379831874743104, "elapsed_seconds": 1.3673947490751743, "env_steps_per_second": 2925.2708500638646, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 7.596257662987337, "eta_seconds": 27346.52758675441, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": -5.544502129484208e-05, "chaser_grad_norm": 0.12448644079267979, "chaser_new_fields": 34.0, "chaser_partner_vision": 0.4230000078678131, "chaser_return": -0.5850003361701965, "chaser_subspace_norm": 0.0030102208256721497, "collisions_per_episode": 2.450000047683716, "entropy": 2.7724441289901733, "explorer_grad_norm": 0.14204901829361916, "explorer_new_fields": 33.525001525878906, "explorer_partner_vision": 0.4230000078678131, "explorer_return": -1.6875, "explorer_subspace_norm": 0.003057338297367096, "final_distance": 4.607369899749756, "max_param_abs": 0.15426939725875854, "policy_loss": -0.0009082589731552737, "value_loss": 3.1435691714286804}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:40.801463+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985420.8014758, "update": 1} +{"device": "cuda", "elapsed_hours": 0.043492414556385486, "elapsed_seconds": 156.57269240298774, "env_steps_per_second": 12773.619520142043, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6962041676990338, "eta_seconds": 6106.335003716522, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0010680055919358722, "chaser_grad_norm": 0.23951933905482292, "chaser_new_fields": 51.79999923706055, "chaser_partner_vision": 0.8662500381469727, "chaser_return": 11.585000038146973, "chaser_subspace_norm": 15.781100273132324, "collisions_per_episode": 10.600000381469727, "entropy": 0.5994224101305008, "explorer_grad_norm": 0.3266453221440315, "explorer_new_fields": 60.5, "explorer_partner_vision": 0.8662500381469727, "explorer_return": 32.07500076293945, "explorer_subspace_norm": 9.417055130004883, "final_distance": 2.0981571674346924, "max_param_abs": 0.3102687895298004, "policy_loss": -0.00553338553896765, "value_loss": 5.397631645202637}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:46:16.006761+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985576.0067735, "update": 500} +{"device": "cuda", "elapsed_hours": 0.09521929496640547, "elapsed_seconds": 342.7894618790597, "env_steps_per_second": 11668.970154663766, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8091666043617038, "eta_seconds": 6512.999775702134, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0010955141924142708, "chaser_grad_norm": 0.34545792639255524, "chaser_new_fields": 51.650001525878906, "chaser_partner_vision": 0.8915000557899475, "chaser_return": 11.6875, "chaser_subspace_norm": 24.363704681396484, "collisions_per_episode": 10.77500057220459, "entropy": 0.42691630870103836, "explorer_grad_norm": 0.39972974359989166, "explorer_new_fields": 62.70000076293945, "explorer_partner_vision": 0.8915000557899475, "explorer_return": 34.72500228881836, "explorer_subspace_norm": 14.47436237335205, "final_distance": 2.1627233028411865, "max_param_abs": 0.36105307936668396, "policy_loss": -0.007009677806877335, "value_loss": 5.382206082344055}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:49:22.223531+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985762.2235453, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.1469803587661267, "elapsed_seconds": 529.1292915580561, "env_steps_per_second": 11339.383579262081, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8127577581155627, "eta_seconds": 6525.927929216025, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0010719877924612042, "chaser_grad_norm": 0.4689277410507202, "chaser_new_fields": 51.79999923706055, "chaser_partner_vision": 0.9125000238418579, "chaser_return": 18.04250144958496, "chaser_subspace_norm": 28.915956497192383, "collisions_per_episode": 16.575000762939453, "entropy": 0.27045851200819016, "explorer_grad_norm": 0.41877812147140503, "explorer_new_fields": 58.95000076293945, "explorer_partner_vision": 0.9125000238418579, "explorer_return": 25.5625, "explorer_subspace_norm": 21.896223068237305, "final_distance": 2.2731075286865234, "max_param_abs": 0.36669817566871643, "policy_loss": -0.004716669003945739, "value_loss": 5.727348208427429}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:52:28.563358+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985948.5633702, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.1988331934516706, "elapsed_seconds": 715.7994964260142, "env_steps_per_second": 11176.314093463307, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7894987410650354, "eta_seconds": 6442.195467834127, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.000877595806509035, "chaser_grad_norm": 0.37057582288980484, "chaser_new_fields": 53.625, "chaser_partner_vision": 0.9142500162124634, "chaser_return": 15.667502403259277, "chaser_subspace_norm": 30.528968811035156, "collisions_per_episode": 14.125, "entropy": 0.2733021527528763, "explorer_grad_norm": 0.30427301675081253, "explorer_new_fields": 64.70000457763672, "explorer_partner_vision": 0.9142500162124634, "explorer_return": 35.525001525878906, "explorer_subspace_norm": 23.450220108032227, "final_distance": 1.869185447692871, "max_param_abs": 0.44789770245552063, "policy_loss": -0.006579066222531793, "value_loss": 5.694641709327698}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:55:35.233564+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986135.2335784, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2525818706916632, "elapsed_seconds": 909.2947344899876, "env_steps_per_second": 10997.534265508399, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7680730948416425, "eta_seconds": 6365.063141429913, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0001859778311383009, "chaser_grad_norm": 0.6821469329297543, "chaser_new_fields": 48.54999923706055, "chaser_partner_vision": 0.8950000405311584, "chaser_return": 16.512500762939453, "chaser_subspace_norm": 32.75629806518555, "collisions_per_episode": 15.824999809265137, "entropy": 0.23705508187413216, "explorer_grad_norm": 0.7226472645998001, "explorer_new_fields": 61.32500076293945, "explorer_partner_vision": 0.8950000405311584, "explorer_return": 28.524999618530273, "explorer_subspace_norm": 25.23515510559082, "final_distance": 1.9461101293563843, "max_param_abs": 0.4811996817588806, "policy_loss": -0.00556038940651149, "value_loss": 5.511416912078857}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:58:48.728803+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986328.7288167, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.30554900060585677, "elapsed_seconds": 1099.9764021810843, "env_steps_per_second": 10909.324942067706, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7314443367665215, "eta_seconds": 6233.199612359477, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0003405500721602728, "chaser_grad_norm": 0.6965800523757935, "chaser_new_fields": 49.97500228881836, "chaser_partner_vision": 0.9337500333786011, "chaser_return": 16.760000228881836, "chaser_subspace_norm": 35.81869125366211, "collisions_per_episode": 15.850000381469727, "entropy": 0.20832566544413567, "explorer_grad_norm": 0.5058369636535645, "explorer_new_fields": 62.35000228881836, "explorer_partner_vision": 0.9337500333786011, "explorer_return": 30.05000114440918, "explorer_subspace_norm": 31.783851623535156, "final_distance": 1.6941133737564087, "max_param_abs": 0.5019632577896118, "policy_loss": -0.0052853168114168625, "value_loss": 5.599910140037537}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:01:59.410470+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986519.4104843, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.35844598529333277, "elapsed_seconds": 1290.4055470559979, "env_steps_per_second": 10849.302401047771, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6898167878114256, "eta_seconds": 6083.340436121132, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0006521052469015132, "chaser_grad_norm": 0.47146280482411385, "chaser_new_fields": 45.400001525878906, "chaser_partner_vision": 0.9045000672340393, "chaser_return": 21.29250144958496, "chaser_subspace_norm": 35.613685607910156, "collisions_per_episode": 20.725000381469727, "entropy": 0.2080659233033657, "explorer_grad_norm": 0.3203674778342247, "explorer_new_fields": 58.625, "explorer_partner_vision": 0.9045000672340393, "explorer_return": 23.0, "explorer_subspace_norm": 33.84025573730469, "final_distance": 1.9768810272216797, "max_param_abs": 0.5346994996070862, "policy_loss": -0.003986125651749717, "value_loss": 5.873714089393616}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:05:09.839616+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986709.8396301, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4113072639752904, "elapsed_seconds": 1480.7061503110453, "env_steps_per_second": 10805.655123833281, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6452290559011615, "eta_seconds": 5922.824601244181, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0007005297155613377, "chaser_grad_norm": 0.4804302081465721, "chaser_new_fields": 46.650001525878906, "chaser_partner_vision": 0.9062500596046448, "chaser_return": 17.98750114440918, "chaser_subspace_norm": 33.837493896484375, "collisions_per_episode": 17.575000762939453, "entropy": 0.21589389443397522, "explorer_grad_norm": 0.37143921107053757, "explorer_new_fields": 59.75, "explorer_partner_vision": 0.9062500596046448, "explorer_return": 24.012500762939453, "explorer_subspace_norm": 32.42313003540039, "final_distance": 2.170441150665283, "max_param_abs": 0.5788803100585938, "policy_loss": -0.005555209225043711, "value_loss": 5.487380385398865}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:08:20.140219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986900.1402342, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.46437382837780106, "elapsed_seconds": 1671.7457821600838, "env_steps_per_second": 10767.18732721549, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5995098533013146, "eta_seconds": 5758.235471884733, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.001144570584860033, "chaser_grad_norm": 0.35064634680747986, "chaser_new_fields": 49.57500076293945, "chaser_partner_vision": 0.9222500324249268, "chaser_return": 18.742502212524414, "chaser_subspace_norm": 34.589786529541016, "collisions_per_episode": 17.975000381469727, "entropy": 0.2092769369482994, "explorer_grad_norm": 0.3287525102496147, "explorer_new_fields": 61.5, "explorer_partner_vision": 0.9222500324249268, "explorer_return": 26.287500381469727, "explorer_subspace_norm": 36.113059997558594, "final_distance": 1.8556467294692993, "max_param_abs": 0.6062168478965759, "policy_loss": -0.005974269750055328, "value_loss": 5.625838398933411}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:11:31.179849+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987091.17986, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5177796121869405, "elapsed_seconds": 1864.0066038729856, "env_steps_per_second": 10729.575720624867, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5533388365608214, "eta_seconds": 5592.019811618957, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0009970206986195107, "chaser_grad_norm": 0.2573762871325016, "chaser_new_fields": 50.22500228881836, "chaser_partner_vision": 0.9145000576972961, "chaser_return": 17.3075008392334, "chaser_subspace_norm": 37.62493896484375, "collisions_per_episode": 16.375, "entropy": 0.20750872045755386, "explorer_grad_norm": 1.036396563053131, "explorer_new_fields": 65.5, "explorer_partner_vision": 0.9145000576972961, "explorer_return": 32.900001525878906, "explorer_subspace_norm": 41.60408020019531, "final_distance": 1.7364215850830078, "max_param_abs": 0.6127142906188965, "policy_loss": -0.005997592179104894, "value_loss": 5.135623574256897}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:14:43.440672+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987283.4406874, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5707457872847186, "elapsed_seconds": 2054.684834224987, "env_steps_per_second": 10707.238226293839, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5046934392051672, "eta_seconds": 5416.896381138602, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00038775515283706596, "chaser_grad_norm": 0.285690575838089, "chaser_new_fields": 50.04999923706055, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 18.220001220703125, "chaser_subspace_norm": 35.49833297729492, "collisions_per_episode": 17.200000762939453, "entropy": 0.1756061539053917, "explorer_grad_norm": 0.520386703312397, "explorer_new_fields": 63.400001525878906, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 29.712499618530273, "explorer_subspace_norm": 44.053672790527344, "final_distance": 1.7026571035385132, "max_param_abs": 0.6178226470947266, "policy_loss": -0.004766876327806813, "value_loss": 5.252135872840881}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:17:54.118901+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987474.1189127, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6225306793069467, "elapsed_seconds": 2241.110445505008, "env_steps_per_second": 10708.97690389903, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4525715850495422, "eta_seconds": 5229.257706178352, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006353142194677197, "chaser_grad_norm": 0.313777856528759, "chaser_new_fields": 51.150001525878906, "chaser_partner_vision": 0.9412500262260437, "chaser_return": 20.9375, "chaser_subspace_norm": 33.63460159301758, "collisions_per_episode": 19.42500114440918, "entropy": 0.20329248532652855, "explorer_grad_norm": 0.3772950693964958, "explorer_new_fields": 61.400001525878906, "explorer_partner_vision": 0.9412500262260437, "explorer_return": 24.399999618530273, "explorer_subspace_norm": 45.88255310058594, "final_distance": 1.4435135126113892, "max_param_abs": 0.6302105188369751, "policy_loss": -0.005886584789493554, "value_loss": 5.6473708152771}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:21:00.544514+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987660.5445278, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6744821378836059, "elapsed_seconds": 2428.1356963809812, "env_steps_per_second": 10707.8035378137, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4008475171428738, "eta_seconds": 5043.051061714345, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006851943465537857, "chaser_grad_norm": 0.6984436810016632, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.9457500576972961, "chaser_return": 20.03000259399414, "chaser_subspace_norm": 32.756927490234375, "collisions_per_episode": 19.0, "entropy": 0.23638136684894562, "explorer_grad_norm": 0.3114742822945118, "explorer_new_fields": 62.150001525878906, "explorer_partner_vision": 0.9457500576972961, "explorer_return": 26.337499618530273, "explorer_subspace_norm": 47.4536018371582, "final_distance": 1.6564123630523682, "max_param_abs": 0.654323935508728, "policy_loss": -0.007503133103252679, "value_loss": 5.643510341644287}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:24:07.569763+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987847.5697749, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.726114080023609, "elapsed_seconds": 2614.0106880849926, "env_steps_per_second": 10711.509378147424, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3484975771867027, "eta_seconds": 4854.5912778721295, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0022721234593765516, "chaser_grad_norm": 0.49326062947511673, "chaser_new_fields": 51.775001525878906, "chaser_partner_vision": 0.9345000386238098, "chaser_return": 19.497501373291016, "chaser_subspace_norm": 39.374732971191406, "collisions_per_episode": 18.075000762939453, "entropy": 0.1542295441031456, "explorer_grad_norm": 0.7382311224937439, "explorer_new_fields": 63.07500076293945, "explorer_partner_vision": 0.9345000386238098, "explorer_return": 27.92500114440918, "explorer_subspace_norm": 48.29940414428711, "final_distance": 1.618445634841919, "max_param_abs": 0.7120906710624695, "policy_loss": -0.005025501692667689, "value_loss": 5.577271223068237}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:27:13.444757+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988033.4447718, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7788387871602834, "elapsed_seconds": 2803.8196337770205, "env_steps_per_second": 10699.689679962421, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.298064645267139, "eta_seconds": 4673.0327229617005, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0010842151161405411, "chaser_grad_norm": 0.5656054019927979, "chaser_new_fields": 48.5, "chaser_partner_vision": 0.8722500205039978, "chaser_return": 17.535001754760742, "chaser_subspace_norm": 36.25432586669922, "collisions_per_episode": 16.899999618530273, "entropy": 0.17374685406684875, "explorer_grad_norm": 0.3251541256904602, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.8722500205039978, "explorer_return": 27.725000381469727, "explorer_subspace_norm": 54.00210952758789, "final_distance": 1.979718804359436, "max_param_abs": 0.7412002086639404, "policy_loss": -0.005915220607072058, "value_loss": 5.703168630599976}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:30:23.253701+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988223.2537117, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8320710270658471, "elapsed_seconds": 2995.45569743705, "env_steps_per_second": 10682.848698907352, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2481065405987708, "eta_seconds": 4493.183546155575, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00020141127725148422, "chaser_grad_norm": 0.5792686082422733, "chaser_new_fields": 47.79999923706055, "chaser_partner_vision": 0.9185000658035278, "chaser_return": 18.597501754760742, "chaser_subspace_norm": 34.14839172363281, "collisions_per_episode": 17.92500114440918, "entropy": 0.20132411643862724, "explorer_grad_norm": 0.819608561694622, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.9185000658035278, "explorer_return": 23.912500381469727, "explorer_subspace_norm": 49.604461669921875, "final_distance": 1.5540332794189453, "max_param_abs": 0.7329506874084473, "policy_loss": -0.004493889589794797, "value_loss": 5.587291955947876}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:33:34.889767+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988414.889782, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8842543893691618, "elapsed_seconds": 3183.3158017289825, "env_steps_per_second": 10680.687094109004, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1963441738523952, "eta_seconds": 4306.839025868623, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005401652732759943, "chaser_grad_norm": 0.6475241631269455, "chaser_new_fields": 49.82500076293945, "chaser_partner_vision": 0.9132500290870667, "chaser_return": 16.160001754760742, "chaser_subspace_norm": 35.741844177246094, "collisions_per_episode": 15.350000381469727, "entropy": 0.22165393829345703, "explorer_grad_norm": 0.2638043612241745, "explorer_new_fields": 61.70000076293945, "explorer_partner_vision": 0.9132500290870667, "explorer_return": 28.5, "explorer_subspace_norm": 51.56485366821289, "final_distance": 2.010725498199463, "max_param_abs": 0.7792694568634033, "policy_loss": -0.005757687604052819, "value_loss": 5.37433922290802}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:36:42.749868+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988602.7498803, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9374293889841647, "elapsed_seconds": 3374.745800342993, "env_steps_per_second": 10667.470123628611, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1457470309806457, "eta_seconds": 4124.689311530325, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0009453642667764939, "chaser_grad_norm": 0.575014017522335, "chaser_new_fields": 43.85000228881836, "chaser_partner_vision": 0.9582500457763672, "chaser_return": 23.252500534057617, "chaser_subspace_norm": 37.42673873901367, "collisions_per_episode": 22.92500114440918, "entropy": 0.20241833478212357, "explorer_grad_norm": 0.30786535143852234, "explorer_new_fields": 58.95000076293945, "explorer_partner_vision": 0.9582500457763672, "explorer_return": 19.387500762939453, "explorer_subspace_norm": 51.406898498535156, "final_distance": 1.7478309869766235, "max_param_abs": 0.8198406100273132, "policy_loss": -0.005340849677104131, "value_loss": 5.7723143100738525}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:39:54.179868+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988794.1798825, "update": 9000} +{"device": "cuda", "elapsed_hours": 0.9899421281428011, "elapsed_seconds": 3563.791661314084, "env_steps_per_second": 10662.800638011533, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0941465626841487, "eta_seconds": 3938.9276256629355, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0010653901419015455, "chaser_grad_norm": 0.7081776857376099, "chaser_new_fields": 48.95000076293945, "chaser_partner_vision": 0.9160000681877136, "chaser_return": 19.297500610351562, "chaser_subspace_norm": 37.90392303466797, "collisions_per_episode": 18.524999618530273, "entropy": 0.18712902069091797, "explorer_grad_norm": 0.6810260117053986, "explorer_new_fields": 59.72500228881836, "explorer_partner_vision": 0.9160000681877136, "explorer_return": 23.537500381469727, "explorer_subspace_norm": 60.93883514404297, "final_distance": 2.08367657661438, "max_param_abs": 0.8115288019180298, "policy_loss": -0.005629435103385205, "value_loss": 5.346372842788696}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:43:03.225728+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988983.2257397, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0428502157969504, "elapsed_seconds": 3754.2607768690214, "env_steps_per_second": 10654.560878256092, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0428502157969504, "eta_seconds": 3754.2607768690214, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0010220249438993578, "chaser_grad_norm": 0.31999313831329346, "chaser_new_fields": 53.17499923706055, "chaser_partner_vision": 0.9432500600814819, "chaser_return": 19.385000228881836, "chaser_subspace_norm": 40.88794708251953, "collisions_per_episode": 17.850000381469727, "entropy": 0.17543289065361023, "explorer_grad_norm": 0.6363183856010437, "explorer_new_fields": 61.22500228881836, "explorer_partner_vision": 0.9432500600814819, "explorer_return": 26.762500762939453, "explorer_subspace_norm": 67.13758850097656, "final_distance": 1.794045329093933, "max_param_abs": 0.8545530438423157, "policy_loss": -0.004681397857146052, "value_loss": 5.027459502220154}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:46:13.694847+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989173.6948624, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.095063632752216, "elapsed_seconds": 3942.2290779079776, "env_steps_per_second": 10653.87098770225, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.990771858204386, "eta_seconds": 3566.7786895357895, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006603466447856476, "chaser_grad_norm": 0.6559595093131065, "chaser_new_fields": 48.10000228881836, "chaser_partner_vision": 0.9450000524520874, "chaser_return": 18.59000015258789, "chaser_subspace_norm": 41.01802062988281, "collisions_per_episode": 17.80000114440918, "entropy": 0.20763841271400452, "explorer_grad_norm": 0.5878511071205139, "explorer_new_fields": 57.150001525878906, "explorer_partner_vision": 0.9450000524520874, "explorer_return": 20.375, "explorer_subspace_norm": 62.97282028198242, "final_distance": 1.844496726989746, "max_param_abs": 0.8928868174552917, "policy_loss": -0.005551177286545084, "value_loss": 5.26336407661438}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:49:21.663145+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989361.6631567, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1467920683569457, "elapsed_seconds": 4128.451446085004, "env_steps_per_second": 10657.749176564748, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9382844195647738, "eta_seconds": 3377.8239104331856, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.000287882837465947, "chaser_grad_norm": 0.1993172951042652, "chaser_new_fields": 51.72500228881836, "chaser_partner_vision": 0.9352500438690186, "chaser_return": 21.079999923706055, "chaser_subspace_norm": 42.487693786621094, "collisions_per_episode": 19.600000381469727, "entropy": 0.1138616893440485, "explorer_grad_norm": 0.25992240011692047, "explorer_new_fields": 63.400001525878906, "explorer_partner_vision": 0.9352500438690186, "explorer_return": 27.387500762939453, "explorer_subspace_norm": 66.75985717773438, "final_distance": 1.694006323814392, "max_param_abs": 0.9298058748245239, "policy_loss": -0.002825849786868062, "value_loss": 4.5673593282699585}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:52:27.885515+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989547.8855288, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.1987139620541711, "elapsed_seconds": 4315.370263395016, "env_steps_per_second": 10659.571993206113, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.886005971953083, "eta_seconds": 3189.6214990310987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0015534461411980116, "chaser_grad_norm": 0.2343033365905285, "chaser_new_fields": 48.57500076293945, "chaser_partner_vision": 0.9325000643730164, "chaser_return": 20.520002365112305, "chaser_subspace_norm": 41.84890365600586, "collisions_per_episode": 19.649999618530273, "entropy": 0.17278479784727097, "explorer_grad_norm": 0.32088904827833176, "explorer_new_fields": 61.95000076293945, "explorer_partner_vision": 0.9325000643730164, "explorer_return": 26.125, "explorer_subspace_norm": 86.00894165039062, "final_distance": 1.7011409997940063, "max_param_abs": 0.8948045372962952, "policy_loss": -0.004950845308137275, "value_loss": 5.235251069068909}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:55:34.804330+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989734.804342, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2503351163555636, "elapsed_seconds": 4501.206418880029, "env_steps_per_second": 10663.807773548666, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8335567442370423, "eta_seconds": 3000.8042792533524, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0009220197721884205, "chaser_grad_norm": 0.24263034760951996, "chaser_new_fields": 50.54999923706055, "chaser_partner_vision": 0.9575000405311584, "chaser_return": 27.74249839782715, "chaser_subspace_norm": 43.15888977050781, "collisions_per_episode": 26.075000762939453, "entropy": 0.1605186127126217, "explorer_grad_norm": 1.6843949258327484, "explorer_new_fields": 59.275001525878906, "explorer_partner_vision": 0.9575000405311584, "explorer_return": 18.037500381469727, "explorer_subspace_norm": 89.84521484375, "final_distance": 1.640454649925232, "max_param_abs": 0.9354342818260193, "policy_loss": -0.004520074210127234, "value_loss": 5.63913094997406}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:58:40.640487+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989920.6405013, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3030432807530612, "elapsed_seconds": 4690.955810711021, "env_steps_per_second": 10658.808570703924, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7818259684518367, "eta_seconds": 2814.5734864266124, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0006457225901885799, "chaser_grad_norm": 0.25561294704675674, "chaser_new_fields": 55.17499923706055, "chaser_partner_vision": 0.956000030040741, "chaser_return": 22.677499771118164, "chaser_subspace_norm": 41.80123519897461, "collisions_per_episode": 20.524999618530273, "entropy": 0.14996552094817162, "explorer_grad_norm": 0.4911758825182915, "explorer_new_fields": 68.9000015258789, "explorer_partner_vision": 0.956000030040741, "explorer_return": 34.3125, "explorer_subspace_norm": 100.54619598388672, "final_distance": 1.7387983798980713, "max_param_abs": 0.9977708458900452, "policy_loss": -0.0038738012369723407, "value_loss": 5.143823266029358}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:01:50.389880+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990110.3898945, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3557565859152885, "elapsed_seconds": 4880.723709295038, "env_steps_per_second": 10654.157681773544, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7300227770313092, "eta_seconds": 2628.081997312713, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0009819166526705558, "chaser_grad_norm": 1.1733236759901047, "chaser_new_fields": 52.67499923706055, "chaser_partner_vision": 0.9545000195503235, "chaser_return": 29.959997177124023, "chaser_subspace_norm": 45.72064971923828, "collisions_per_episode": 27.950000762939453, "entropy": 0.15483557805418968, "explorer_grad_norm": 0.5292600095272064, "explorer_new_fields": 61.35000228881836, "explorer_partner_vision": 0.9545000195503235, "explorer_return": 18.0, "explorer_subspace_norm": 73.9805908203125, "final_distance": 1.5375436544418335, "max_param_abs": 1.0113674402236938, "policy_loss": -0.00482884364854641, "value_loss": 5.029845356941223}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:05:00.157778+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990300.1577923, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.408538630305007, "elapsed_seconds": 5070.739069098025, "env_steps_per_second": 10649.335188451225, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6781852664431515, "eta_seconds": 2441.466959195345, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0005181762678354573, "chaser_grad_norm": 0.31137532368302345, "chaser_new_fields": 46.75, "chaser_partner_vision": 0.9380000233650208, "chaser_return": 19.41750144958496, "chaser_subspace_norm": 39.6579475402832, "collisions_per_episode": 18.875, "entropy": 0.2123427726328373, "explorer_grad_norm": 0.26607783138751984, "explorer_new_fields": 57.22500228881836, "explorer_partner_vision": 0.9380000233650208, "explorer_return": 19.462499618530273, "explorer_subspace_norm": 79.50724029541016, "final_distance": 1.6376034021377563, "max_param_abs": 1.013170838356018, "policy_loss": -0.004818619448691663, "value_loss": 5.378113269805908}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:08:10.173137+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990490.1731474, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4611226177339105, "elapsed_seconds": 5260.041423842078, "env_steps_per_second": 10646.303990339313, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6261954076002475, "eta_seconds": 2254.303467360891, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -2.5735723701458824e-05, "chaser_grad_norm": 1.2067946642637253, "chaser_new_fields": 51.17499923706055, "chaser_partner_vision": 0.9545000195503235, "chaser_return": 21.114999771118164, "chaser_subspace_norm": 50.2652587890625, "collisions_per_episode": 19.75, "entropy": 0.13857953995466232, "explorer_grad_norm": 0.60109031945467, "explorer_new_fields": 61.60000228881836, "explorer_partner_vision": 0.9545000195503235, "explorer_return": 25.73750114440918, "explorer_subspace_norm": 80.0230484008789, "final_distance": 1.5739116668701172, "max_param_abs": 1.0376462936401367, "policy_loss": -0.003016507959920034, "value_loss": 5.500334024429321}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:11:19.475492+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990679.4755077, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.515197290595, "elapsed_seconds": 5454.710246142, "env_steps_per_second": 10633.012090976265, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5747300067774137, "eta_seconds": 2069.0280243986895, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0010817054275791271, "chaser_grad_norm": 0.253335814923048, "chaser_new_fields": 51.875, "chaser_partner_vision": 0.9505000710487366, "chaser_return": 21.125, "chaser_subspace_norm": 46.05088806152344, "collisions_per_episode": 19.600000381469727, "entropy": 0.15126802027225494, "explorer_grad_norm": 0.6774330735206604, "explorer_new_fields": 59.97500228881836, "explorer_partner_vision": 0.9505000710487366, "explorer_return": 23.375, "explorer_subspace_norm": 90.76144409179688, "final_distance": 1.6682811975479126, "max_param_abs": 1.1371933221817017, "policy_loss": -0.0040234598433599444, "value_loss": 5.27064836025238}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:14:34.144314+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990874.1443245, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5691623572805677, "elapsed_seconds": 5648.9844862100435, "env_steps_per_second": 10621.37808069191, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5230541190935225, "eta_seconds": 1882.994828736681, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0001613329575222977, "chaser_grad_norm": 0.19970540329813957, "chaser_new_fields": 47.875, "chaser_partner_vision": 0.9565000534057617, "chaser_return": 18.625, "chaser_subspace_norm": 44.59516525268555, "collisions_per_episode": 17.850000381469727, "entropy": 0.14151310175657272, "explorer_grad_norm": 0.42165037617087364, "explorer_new_fields": 61.47500228881836, "explorer_partner_vision": 0.9565000534057617, "explorer_return": 25.524999618530273, "explorer_subspace_norm": 103.01426696777344, "final_distance": 2.021512508392334, "max_param_abs": 1.1552108526229858, "policy_loss": -0.0032181050330351013, "value_loss": 4.4845792055130005}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:17:48.418555+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991068.4185684, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6225639346386176, "elapsed_seconds": 5841.230164699024, "env_steps_per_second": 10614.202531290706, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4710669487660502, "eta_seconds": 1695.8410155577808, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": -0.00012916255756995065, "chaser_grad_norm": 0.1414089798927307, "chaser_new_fields": 52.900001525878906, "chaser_partner_vision": 0.9507500529289246, "chaser_return": 20.850000381469727, "chaser_subspace_norm": 46.136619567871094, "collisions_per_episode": 18.950000762939453, "entropy": 0.1253863200545311, "explorer_grad_norm": 0.6366455852985382, "explorer_new_fields": 62.60000228881836, "explorer_partner_vision": 0.9507500529289246, "explorer_return": 26.700000762939453, "explorer_subspace_norm": 107.07857513427734, "final_distance": 1.5671285390853882, "max_param_abs": 1.1938846111297607, "policy_loss": -0.002721733722138353, "value_loss": 4.751402974128723}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:21:00.664233+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991260.664247, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6764966594797444, "elapsed_seconds": 6035.387974127079, "env_steps_per_second": 10604.123591450898, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4191241648699361, "eta_seconds": 1508.8469935317698, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0007563322123180782, "chaser_grad_norm": 0.31364617869257927, "chaser_new_fields": 53.650001525878906, "chaser_partner_vision": 0.9385000467300415, "chaser_return": 26.912500381469727, "chaser_subspace_norm": 46.063045501708984, "collisions_per_episode": 24.475000381469727, "entropy": 0.14937692508101463, "explorer_grad_norm": 0.351222887635231, "explorer_new_fields": 63.5, "explorer_partner_vision": 0.9385000467300415, "explorer_return": 25.287500381469727, "explorer_subspace_norm": 91.6572036743164, "final_distance": 1.5849348306655884, "max_param_abs": 1.375173568725586, "policy_loss": -0.0036606949469994987, "value_loss": 4.9808595180511475}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:24:14.822041+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991454.822055, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7295621051947172, "elapsed_seconds": 6226.423578700982, "env_steps_per_second": 10599.985556037222, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.36687681019281887, "eta_seconds": 1320.7565166941479, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005990040566423716, "chaser_grad_norm": 0.32785581797361374, "chaser_new_fields": 44.32500076293945, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 33.467498779296875, "chaser_subspace_norm": 43.56680679321289, "collisions_per_episode": 32.125, "entropy": 0.17596200853586197, "explorer_grad_norm": 0.6169940456748009, "explorer_new_fields": 53.125, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 7.625, "explorer_subspace_norm": 101.58419799804688, "final_distance": 1.3526531457901, "max_param_abs": 1.407278299331665, "policy_loss": -0.0033193524461250234, "value_loss": 6.302483677864075}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:27:25.857645+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991645.857656, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7813117908130516, "elapsed_seconds": 6412.722446926986, "env_steps_per_second": 10603.920653479398, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3143491395552444, "eta_seconds": 1131.6569023988798, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0005020015431549951, "chaser_grad_norm": 0.4954511448740959, "chaser_new_fields": 51.0, "chaser_partner_vision": 0.9327500462532043, "chaser_return": 19.91750144958496, "chaser_subspace_norm": 43.982460021972656, "collisions_per_episode": 18.625, "entropy": 0.19081757590174675, "explorer_grad_norm": 0.4349144473671913, "explorer_new_fields": 60.35000228881836, "explorer_partner_vision": 0.9327500462532043, "explorer_return": 23.9375, "explorer_subspace_norm": 103.49979400634766, "final_distance": 2.075138568878174, "max_param_abs": 1.5087449550628662, "policy_loss": -0.004728602130898096, "value_loss": 5.160322189331055}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:30:32.156515+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991832.1565294, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8338382688497141, "elapsed_seconds": 6601.817767858971, "env_steps_per_second": 10603.140295813047, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2619768955499592, "eta_seconds": 943.1168239798529, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.000571412075923039, "chaser_grad_norm": 0.9630060940980911, "chaser_new_fields": 48.67499923706055, "chaser_partner_vision": 0.9407500624656677, "chaser_return": 25.670001983642578, "chaser_subspace_norm": 41.90203857421875, "collisions_per_episode": 24.200000762939453, "entropy": 0.17032190412282944, "explorer_grad_norm": 0.4771170914173126, "explorer_new_fields": 58.95000076293945, "explorer_partner_vision": 0.9407500624656677, "explorer_return": 18.787500381469727, "explorer_subspace_norm": 94.86479949951172, "final_distance": 1.7401539087295532, "max_param_abs": 1.5417625904083252, "policy_loss": -0.003824234211815991, "value_loss": 5.3143463134765625}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:33:41.251835+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992021.2518477, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.8876398450597318, "elapsed_seconds": 6795.503442215035, "env_steps_per_second": 10595.24148758744, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.20973776056219245, "eta_seconds": 755.0559380238928, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 8.921171379405735e-05, "chaser_grad_norm": 0.6467003673315048, "chaser_new_fields": 47.29999923706055, "chaser_partner_vision": 0.9547500610351562, "chaser_return": 23.5, "chaser_subspace_norm": 45.374027252197266, "collisions_per_episode": 22.600000381469727, "entropy": 0.19363467767834663, "explorer_grad_norm": 0.29173319786787033, "explorer_new_fields": 56.97500228881836, "explorer_partner_vision": 0.9547500610351562, "explorer_return": 17.5625, "explorer_subspace_norm": 94.13497161865234, "final_distance": 1.6043630838394165, "max_param_abs": 1.5841625928878784, "policy_loss": -0.005353930544110597, "value_loss": 5.722419023513794}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:36:54.937510+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992214.9375253, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9405314423402888, "elapsed_seconds": 6985.91319242504, "env_steps_per_second": 10592.745423782195, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.15734038721678015, "eta_seconds": 566.4253939804086, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005189570293695489, "chaser_grad_norm": 0.848080575466156, "chaser_new_fields": 47.57500076293945, "chaser_partner_vision": 0.9520000219345093, "chaser_return": 19.510000228881836, "chaser_subspace_norm": 44.41987228393555, "collisions_per_episode": 18.80000114440918, "entropy": 0.18391158804297447, "explorer_grad_norm": 0.19431817904114723, "explorer_new_fields": 57.45000076293945, "explorer_partner_vision": 0.9520000219345093, "explorer_return": 21.899999618530273, "explorer_subspace_norm": 110.77628326416016, "final_distance": 1.6811773777008057, "max_param_abs": 1.5921014547348022, "policy_loss": -0.003955657614859831, "value_loss": 5.387449741363525}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:40:05.347259+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992405.3472714, "update": 18500} +{"device": "cuda", "elapsed_hours": 1.9938400112766734, "elapsed_seconds": 7177.824040596024, "env_steps_per_second": 10588.167050371048, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10493894796193017, "eta_seconds": 377.7802126629486, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005451329075158995, "chaser_grad_norm": 0.37418269738554955, "chaser_new_fields": 50.04999923706055, "chaser_partner_vision": 0.940500020980835, "chaser_return": 18.954999923706055, "chaser_subspace_norm": 44.560760498046875, "collisions_per_episode": 17.950000762939453, "entropy": 0.19704873487353325, "explorer_grad_norm": 0.42180851101875305, "explorer_new_fields": 59.85000228881836, "explorer_partner_vision": 0.940500020980835, "explorer_return": 24.98750114440918, "explorer_subspace_norm": 102.77297973632812, "final_distance": 1.5011918544769287, "max_param_abs": 1.5925586223602295, "policy_loss": -0.004980837227776647, "value_loss": 4.927888989448547}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:43:17.258109+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992597.2581236, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.047485760755565, "elapsed_seconds": 7370.948738720035, "env_steps_per_second": 10582.084174627525, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05249963489116834, "eta_seconds": 188.99868560820602, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0008904130132443289, "chaser_grad_norm": 0.37351489439606667, "chaser_new_fields": 49.125, "chaser_partner_vision": 0.9467500448226929, "chaser_return": 19.60000228881836, "chaser_subspace_norm": 47.347591400146484, "collisions_per_episode": 18.700000762939453, "entropy": 0.1791943944990635, "explorer_grad_norm": 1.401249885559082, "explorer_new_fields": 58.400001525878906, "explorer_partner_vision": 0.9467500448226929, "explorer_return": 21.350000381469727, "explorer_subspace_norm": 141.62266540527344, "final_distance": 1.583022117614746, "max_param_abs": 1.6378453969955444, "policy_loss": -0.004394496601541675, "value_loss": 5.470173954963684}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:46:30.382808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992790.382822, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.0996697709600105, "elapsed_seconds": 7558.811175456038, "env_steps_per_second": 10583.674885247208, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:49:38.245242+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992978.2452521, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.099672739202215, "elapsed_seconds": 7558.8218611279735, "env_steps_per_second": 10583.659923434405, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.00025015939701367174, "chaser_grad_norm": 0.31481391191482544, "chaser_new_fields": 49.900001525878906, "chaser_partner_vision": 0.9252500534057617, "chaser_return": 21.020002365112305, "chaser_subspace_norm": 49.21955108642578, "collisions_per_episode": 19.80000114440918, "entropy": 0.1560382954776287, "explorer_grad_norm": 0.6951030343770981, "explorer_new_fields": 59.42499923706055, "explorer_partner_vision": 0.9252500534057617, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 113.72798919677734, "final_distance": 1.6350990533828735, "max_param_abs": 1.6354095935821533, "policy_loss": -0.0035889741121788443, "value_loss": 4.686129331588745}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:49:38.255928+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992978.2559416, "update": 20000} diff --git a/social/seed_0006/attempt_01/status.json b/social/seed_0006/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..0162da113d6c7d16d055f10dbd34af9ac41f16b5 --- /dev/null +++ b/social/seed_0006/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.099672739202215, + "elapsed_seconds": 7558.8218611279735, + "env_steps_per_second": 10583.659923434405, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.00025015939701367174, + "chaser_grad_norm": 0.31481391191482544, + "chaser_new_fields": 49.900001525878906, + "chaser_partner_vision": 0.9252500534057617, + "chaser_return": 21.020002365112305, + "chaser_subspace_norm": 49.21955108642578, + "collisions_per_episode": 19.80000114440918, + "entropy": 0.1560382954776287, + "explorer_grad_norm": 0.6951030343770981, + "explorer_new_fields": 59.42499923706055, + "explorer_partner_vision": 0.9252500534057617, + "explorer_return": 22.48750114440918, + "explorer_subspace_norm": 113.72798919677734, + "final_distance": 1.6350990533828735, + "max_param_abs": 1.6354095935821533, + "policy_loss": -0.0035889741121788443, + "value_loss": 4.686129331588745 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:49:38.255928+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782992978.2559416, + "update": 20000 +} diff --git a/social/seed_0007/attempt_01/checkpoints/latest.safetensors b/social/seed_0007/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..31326416f74a6429fee49e256671f1c765e0498d --- /dev/null +++ b/social/seed_0007/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:cb8a8e1bcbbb446b3ce500375218bbb11f70265bebe4d228299de67891943d15 +size 11460408 diff --git a/social/seed_0007/attempt_01/config.json b/social/seed_0007/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..a1d3144739bbd8b9560531b482308e5d3c7a66ab --- /dev/null +++ b/social/seed_0007/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0007.attempt_01", + "seed": 7, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0007/attempt_01/metrics.jsonl b/social/seed_0007/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..78b54418a8f0094fd4e5cc7ea40d6912879ce3b5 --- /dev/null +++ b/social/seed_0007/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.00040586303748164535, "elapsed_seconds": 1.4611069349339232, "env_steps_per_second": 2737.6504103588386, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 8.116854886595426, "eta_seconds": 29220.677591743533, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 1.7225777219215255e-05, "chaser_grad_norm": 0.2306205853819847, "chaser_new_fields": 34.07500076293945, "chaser_partner_vision": 0.33925002813339233, "chaser_return": -1.7475003004074097, "chaser_subspace_norm": 0.003010429907590151, "collisions_per_episode": 1.375, "entropy": 2.772493600845337, "explorer_grad_norm": 0.12197021394968033, "explorer_new_fields": 34.67499923706055, "explorer_partner_vision": 0.33925002813339233, "explorer_return": 0.987500011920929, "explorer_subspace_norm": 0.002479015849530697, "final_distance": 5.210450649261475, "max_param_abs": 0.1542869657278061, "policy_loss": -0.0007321000248197151, "value_loss": 3.1650837063789368}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:43:50.914491+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985430.9145014, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05001195871164479, "elapsed_seconds": 180.04305136192124, "env_steps_per_second": 11108.454255085993, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9504663897541468, "eta_seconds": 7021.6790031149285, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0012214580853706636, "chaser_grad_norm": 0.38530056923627853, "chaser_new_fields": 60.17499923706055, "chaser_partner_vision": 0.5035000443458557, "chaser_return": 10.609999656677246, "chaser_subspace_norm": 13.120227813720703, "collisions_per_episode": 8.050000190734863, "entropy": 0.7936360985040665, "explorer_grad_norm": 0.3370715156197548, "explorer_new_fields": 63.47500228881836, "explorer_partner_vision": 0.5035000443458557, "explorer_return": 38.712501525878906, "explorer_subspace_norm": 6.2786078453063965, "final_distance": 3.5043983459472656, "max_param_abs": 0.3146236538887024, "policy_loss": -0.006486020108219748, "value_loss": 5.4466928243637085}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:46:49.496437+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985609.4964495, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10343476464415693, "elapsed_seconds": 372.36515271896496, "env_steps_per_second": 10742.143755376914, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9652605282389815, "eta_seconds": 7074.937901660333, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 5.608291446712599e-05, "chaser_grad_norm": 0.24521180987358093, "chaser_new_fields": 58.04999923706055, "chaser_partner_vision": 0.5387500524520874, "chaser_return": 13.357500076293945, "chaser_subspace_norm": 25.77691078186035, "collisions_per_episode": 11.02500057220459, "entropy": 0.5038723647594452, "explorer_grad_norm": 0.4792740195989609, "explorer_new_fields": 62.150001525878906, "explorer_partner_vision": 0.5387500524520874, "explorer_return": 34.97500228881836, "explorer_subspace_norm": 8.947600364685059, "final_distance": 3.8781654834747314, "max_param_abs": 0.3736887574195862, "policy_loss": -0.006218448474072602, "value_loss": 6.056857585906982}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:50:01.818538+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985801.818553, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15513081739750406, "elapsed_seconds": 558.4709426310146, "env_steps_per_second": 10743.620736529956, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9132800812358834, "eta_seconds": 6887.80829244918, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00021027151542241107, "chaser_grad_norm": 0.29203951358795166, "chaser_new_fields": 55.0, "chaser_partner_vision": 0.5607500076293945, "chaser_return": 14.614999771118164, "chaser_subspace_norm": 34.197410583496094, "collisions_per_episode": 12.699999809265137, "entropy": 0.44512469321489334, "explorer_grad_norm": 0.38833121955394745, "explorer_new_fields": 62.07500076293945, "explorer_partner_vision": 0.5607500076293945, "explorer_return": 33.53750228881836, "explorer_subspace_norm": 9.94297981262207, "final_distance": 3.6187844276428223, "max_param_abs": 0.431519478559494, "policy_loss": -0.004219361965553503, "value_loss": 5.964339375495911}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:07.924327+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985987.9243376, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.20673872466137658, "elapsed_seconds": 744.2594087809557, "env_steps_per_second": 10748.940363553394, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8606485219523892, "eta_seconds": 6698.334679028601, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 8.79957391637376e-05, "chaser_grad_norm": 0.2488110437989235, "chaser_new_fields": 50.47500228881836, "chaser_partner_vision": 0.6885000467300415, "chaser_return": 13.917500495910645, "chaser_subspace_norm": 37.1536979675293, "collisions_per_episode": 13.02500057220459, "entropy": 0.3740861937403679, "explorer_grad_norm": 0.22383227199316025, "explorer_new_fields": 61.625, "explorer_partner_vision": 0.6885000467300415, "explorer_return": 31.725000381469727, "explorer_subspace_norm": 12.465659141540527, "final_distance": 2.930763006210327, "max_param_abs": 0.4778977930545807, "policy_loss": -0.006696206575256269, "value_loss": 5.8274126052856445}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:13.712794+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986173.7128065, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2589220674227626, "elapsed_seconds": 932.1194427219452, "env_steps_per_second": 10728.238830421047, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.812454471959338, "eta_seconds": 6524.836099053617, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002934313613573991, "chaser_grad_norm": 0.5777267143130302, "chaser_new_fields": 49.125, "chaser_partner_vision": 0.7642500400543213, "chaser_return": 13.665000915527344, "chaser_subspace_norm": 39.24969482421875, "collisions_per_episode": 13.050000190734863, "entropy": 0.33152108639478683, "explorer_grad_norm": 0.33416763693094254, "explorer_new_fields": 60.07500076293945, "explorer_partner_vision": 0.7642500400543213, "explorer_return": 29.6875, "explorer_subspace_norm": 15.869560241699219, "final_distance": 2.3920845985412598, "max_param_abs": 0.5382234454154968, "policy_loss": -0.006602799023504247, "value_loss": 5.747146248817444}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:21.572829+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986361.5728433, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3121380988905568, "elapsed_seconds": 1123.6971560060047, "env_steps_per_second": 10679.033880135474, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7687825603798222, "eta_seconds": 6367.617217367359, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": -0.00013435416840934522, "chaser_grad_norm": 0.5557942278683186, "chaser_new_fields": 50.875, "chaser_partner_vision": 0.8740000128746033, "chaser_return": 17.31999969482422, "chaser_subspace_norm": 39.82155227661133, "collisions_per_episode": 16.399999618530273, "entropy": 0.2341924123466015, "explorer_grad_norm": 0.32192814350128174, "explorer_new_fields": 62.85000228881836, "explorer_partner_vision": 0.8740000128746033, "explorer_return": 29.8125, "explorer_subspace_norm": 19.654090881347656, "final_distance": 1.8405544757843018, "max_param_abs": 0.5431714653968811, "policy_loss": -0.0051984627335386335, "value_loss": 5.487710356712341}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:02:33.150541+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986553.1505551, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.36505343838138893, "elapsed_seconds": 1314.1923781730002, "env_steps_per_second": 10652.93044802383, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.720966209512262, "eta_seconds": 6195.478354244143, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.000842883990720189, "chaser_grad_norm": 0.4221755713224411, "chaser_new_fields": 49.775001525878906, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 17.10500144958496, "chaser_subspace_norm": 35.06961441040039, "collisions_per_episode": 16.30000114440918, "entropy": 0.23597263917326927, "explorer_grad_norm": 0.3897154852747917, "explorer_new_fields": 61.54999923706055, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 29.11250114440918, "explorer_subspace_norm": 22.740652084350586, "final_distance": 1.912966012954712, "max_param_abs": 0.5908340811729431, "policy_loss": -0.005241284899412335, "value_loss": 5.1882244348526}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:05:43.645762+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986743.6457725, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.4192082702677645, "elapsed_seconds": 1509.1497729639523, "env_steps_per_second": 10601.996095176286, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.676833081071058, "eta_seconds": 6036.599091855809, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": -0.0001944849833361295, "chaser_grad_norm": 0.3281169980764389, "chaser_new_fields": 49.04999923706055, "chaser_partner_vision": 0.9110000729560852, "chaser_return": 17.982501983642578, "chaser_subspace_norm": 37.71406555175781, "collisions_per_episode": 17.274999618530273, "entropy": 0.25503765046596527, "explorer_grad_norm": 0.23140599206089973, "explorer_new_fields": 60.17499923706055, "explorer_partner_vision": 0.9110000729560852, "explorer_return": 25.36250114440918, "explorer_subspace_norm": 21.39800262451172, "final_distance": 2.0913333892822266, "max_param_abs": 0.6075982451438904, "policy_loss": -0.005121774894192299, "value_loss": 5.74100649356842}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:08:58.603157+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986938.6031709, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4723296610252712, "elapsed_seconds": 1700.3867796909763, "env_steps_per_second": 10585.826833628566, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.626913276864823, "eta_seconds": 5856.887796713362, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0012833841207253727, "chaser_grad_norm": 0.38843734189867973, "chaser_new_fields": 49.70000076293945, "chaser_partner_vision": 0.9395000338554382, "chaser_return": 18.45250129699707, "chaser_subspace_norm": 38.411312103271484, "collisions_per_episode": 17.375, "entropy": 0.2406049370765686, "explorer_grad_norm": 0.3173113092780113, "explorer_new_fields": 63.275001525878906, "explorer_partner_vision": 0.9395000338554382, "explorer_return": 29.662500381469727, "explorer_subspace_norm": 25.611873626708984, "final_distance": 1.7626034021377563, "max_param_abs": 0.6345022916793823, "policy_loss": -0.00560725985746835, "value_loss": 5.377776622772217}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:09.840163+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987129.8401744, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5266681925841517, "elapsed_seconds": 1896.005493302946, "env_steps_per_second": 10548.492644480106, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.580004577752455, "eta_seconds": 5688.016479908838, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0003196362783325357, "chaser_grad_norm": 0.33714716881513596, "chaser_new_fields": 45.79999923706055, "chaser_partner_vision": 0.9282500147819519, "chaser_return": 17.680002212524414, "chaser_subspace_norm": 36.87439727783203, "collisions_per_episode": 17.55000114440918, "entropy": 0.26130300015211105, "explorer_grad_norm": 0.3457140661776066, "explorer_new_fields": 58.22500228881836, "explorer_partner_vision": 0.9282500147819519, "explorer_return": 22.5625, "explorer_subspace_norm": 28.602895736694336, "final_distance": 1.7747173309326172, "max_param_abs": 0.6275982856750488, "policy_loss": -0.006178845185786486, "value_loss": 5.6645355224609375}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:25.458878+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987325.458892, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5799648466191462, "elapsed_seconds": 2087.873447828926, "env_steps_per_second": 10537.03710963741, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5289982319959308, "eta_seconds": 5504.393635185351, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0003872640844075903, "chaser_grad_norm": 0.17515183240175247, "chaser_new_fields": 51.95000076293945, "chaser_partner_vision": 0.9452500343322754, "chaser_return": 16.33500099182129, "chaser_subspace_norm": 38.50767517089844, "collisions_per_episode": 15.15000057220459, "entropy": 0.16464586928486824, "explorer_grad_norm": 0.26043545082211494, "explorer_new_fields": 64.7750015258789, "explorer_partner_vision": 0.9452500343322754, "explorer_return": 32.125, "explorer_subspace_norm": 36.97144317626953, "final_distance": 1.7047550678253174, "max_param_abs": 0.6525636911392212, "policy_loss": -0.004116567297373308, "value_loss": 5.231220245361328}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:18:37.326834+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987517.3268473, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6330253719969187, "elapsed_seconds": 2278.8913391889073, "env_steps_per_second": 10531.43675053062, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4770592013261437, "eta_seconds": 5317.413124774117, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0005908832708367839, "chaser_grad_norm": 0.2411206215620041, "chaser_new_fields": 49.45000076293945, "chaser_partner_vision": 0.9195000529289246, "chaser_return": 18.387502670288086, "chaser_subspace_norm": 40.77865982055664, "collisions_per_episode": 17.42500114440918, "entropy": 0.2173367701470852, "explorer_grad_norm": 0.4288967475295067, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.9195000529289246, "explorer_return": 25.700000762939453, "explorer_subspace_norm": 35.04820251464844, "final_distance": 1.9826183319091797, "max_param_abs": 0.6536448001861572, "policy_loss": -0.005588211885187988, "value_loss": 5.880403757095337}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:21:48.344725+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987708.3447404, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6870898525630602, "elapsed_seconds": 2473.523469227017, "env_steps_per_second": 10511.321329053359, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4270327707078942, "eta_seconds": 5137.317974548419, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00041278279421130826, "chaser_grad_norm": 0.3685663118958473, "chaser_new_fields": 47.10000228881836, "chaser_partner_vision": 0.9407500624656677, "chaser_return": 18.492502212524414, "chaser_subspace_norm": 37.52228546142578, "collisions_per_episode": 18.075000762939453, "entropy": 0.23411992564797401, "explorer_grad_norm": 0.4071793183684349, "explorer_new_fields": 57.900001525878906, "explorer_partner_vision": 0.9407500624656677, "explorer_return": 20.98750114440918, "explorer_subspace_norm": 35.51255798339844, "final_distance": 1.8107002973556519, "max_param_abs": 0.7019726634025574, "policy_loss": -0.006607039872559284, "value_loss": 5.634939193725586}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:25:02.976855+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987902.9768689, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7394879773727735, "elapsed_seconds": 2662.1567185419844, "env_steps_per_second": 10517.788004357271, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3733348151208649, "eta_seconds": 4944.005334435114, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0006493375085945274, "chaser_grad_norm": 0.6265990249812603, "chaser_new_fields": 50.625, "chaser_partner_vision": 0.9542500376701355, "chaser_return": 21.002500534057617, "chaser_subspace_norm": 41.17892837524414, "collisions_per_episode": 19.625, "entropy": 0.16287459060549736, "explorer_grad_norm": 0.3860454633831978, "explorer_new_fields": 60.29999923706055, "explorer_partner_vision": 0.9542500376701355, "explorer_return": 24.48750114440918, "explorer_subspace_norm": 39.90684509277344, "final_distance": 1.6728992462158203, "max_param_abs": 0.7221271395683289, "policy_loss": -0.003207497250755864, "value_loss": 5.471356153488159}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:28:11.610103+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988091.6101172, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.792913328756921, "elapsed_seconds": 2854.4879835249158, "env_steps_per_second": 10509.765734923138, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3215222145948684, "eta_seconds": 4757.479972541526, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0005193300298285131, "chaser_grad_norm": 0.6283604726195335, "chaser_new_fields": 51.29999923706055, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 19.450000762939453, "chaser_subspace_norm": 43.48054122924805, "collisions_per_episode": 18.200000762939453, "entropy": 0.14234085753560066, "explorer_grad_norm": 0.4614320918917656, "explorer_new_fields": 66.4749984741211, "explorer_partner_vision": 0.9455000162124634, "explorer_return": 34.462501525878906, "explorer_subspace_norm": 46.20787048339844, "final_distance": 1.8070316314697266, "max_param_abs": 0.7578325867652893, "policy_loss": -0.004141254453452703, "value_loss": 5.054668426513672}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:31:23.941369+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988283.9413822, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8464847681605412, "elapsed_seconds": 3047.3451653779484, "env_steps_per_second": 10500.943694716376, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2697271522408118, "eta_seconds": 4571.017748066923, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0007379790333299496, "chaser_grad_norm": 0.39755718410015106, "chaser_new_fields": 48.75, "chaser_partner_vision": 0.9217500686645508, "chaser_return": 19.99250030517578, "chaser_subspace_norm": 43.474609375, "collisions_per_episode": 19.075000762939453, "entropy": 0.1508861966431141, "explorer_grad_norm": 0.3449355848133564, "explorer_new_fields": 62.17499923706055, "explorer_partner_vision": 0.9217500686645508, "explorer_return": 26.337499618530273, "explorer_subspace_norm": 43.05052947998047, "final_distance": 1.771030068397522, "max_param_abs": 0.7631932497024536, "policy_loss": -0.003999835048802414, "value_loss": 5.323531627655029}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:34:36.798550+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988476.7985654, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.8991473324197189, "elapsed_seconds": 3236.930396710988, "env_steps_per_second": 10503.778528740393, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2164934497443256, "eta_seconds": 4379.376419079572, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0009742745987870727, "chaser_grad_norm": 0.5592991709709167, "chaser_new_fields": 48.57500076293945, "chaser_partner_vision": 0.952750027179718, "chaser_return": 18.985000610351562, "chaser_subspace_norm": 42.25828552246094, "collisions_per_episode": 18.350000381469727, "entropy": 0.20311175286769867, "explorer_grad_norm": 0.2718800753355026, "explorer_new_fields": 61.29999923706055, "explorer_partner_vision": 0.952750027179718, "explorer_return": 25.387500762939453, "explorer_subspace_norm": 42.85911178588867, "final_distance": 1.7557395696640015, "max_param_abs": 0.784469485282898, "policy_loss": -0.005660233523698821, "value_loss": 5.5022361278533936}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:37:46.383780+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988666.3837907, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9525022372758314, "elapsed_seconds": 3429.008054192993, "env_steps_per_second": 10498.663004299211, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.164169401114905, "eta_seconds": 4191.009844013658, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005964448344597528, "chaser_grad_norm": 0.3135776072740555, "chaser_new_fields": 51.5, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 16.872501373291016, "chaser_subspace_norm": 40.81077194213867, "collisions_per_episode": 15.725000381469727, "entropy": 0.20348817110061646, "explorer_grad_norm": 0.983067974448204, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 28.80000114440918, "explorer_subspace_norm": 47.05223083496094, "final_distance": 1.6365560293197632, "max_param_abs": 0.7737571001052856, "policy_loss": -0.005564030335747816, "value_loss": 5.098546624183655}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:40:58.461439+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988858.4614546, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0042954465177738, "elapsed_seconds": 3615.4636074639857, "env_steps_per_second": 10510.408657288228, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1100107566775395, "eta_seconds": 3996.038724039142, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0009518171375211271, "chaser_grad_norm": 0.5983955450356007, "chaser_new_fields": 46.22500228881836, "chaser_partner_vision": 0.9137500524520874, "chaser_return": 22.892501831054688, "chaser_subspace_norm": 45.662044525146484, "collisions_per_episode": 22.274999618530273, "entropy": 0.15417281910777092, "explorer_grad_norm": 0.33269526809453964, "explorer_new_fields": 57.35000228881836, "explorer_partner_vision": 0.9137500524520874, "explorer_return": 18.887500762939453, "explorer_subspace_norm": 52.06460952758789, "final_distance": 1.8156722784042358, "max_param_abs": 0.8136343955993652, "policy_loss": -0.0035538669140038337, "value_loss": 5.842320203781128}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:44:04.916991+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989044.9170039, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0577723826661047, "elapsed_seconds": 3807.9805775979767, "env_steps_per_second": 10504.255256793213, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0577723826661047, "eta_seconds": 3807.9805775979767, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -0.00034085107959747063, "chaser_grad_norm": 0.2753322198987007, "chaser_new_fields": 52.20000076293945, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 18.38249969482422, "chaser_subspace_norm": 46.327903747558594, "collisions_per_episode": 17.075000762939453, "entropy": 0.09131265804171562, "explorer_grad_norm": 0.8011827766895294, "explorer_new_fields": 64.9000015258789, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 32.36249923706055, "explorer_subspace_norm": 61.146175384521484, "final_distance": 1.551917314529419, "max_param_abs": 0.8219314217567444, "policy_loss": -0.0016369569194552014, "value_loss": 5.039229869842529}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:47:17.433962+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989237.4339752, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1112709329544386, "elapsed_seconds": 4000.575358635979, "env_steps_per_second": 10498.48990079271, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.005435606006397, "eta_seconds": 3619.5681816230285, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0018799541420764232, "chaser_grad_norm": 0.5184521190822124, "chaser_new_fields": 43.82500076293945, "chaser_partner_vision": 0.9505000710487366, "chaser_return": 20.297500610351562, "chaser_subspace_norm": 44.82428741455078, "collisions_per_episode": 20.274999618530273, "entropy": 0.1705855019390583, "explorer_grad_norm": 0.495189793407917, "explorer_new_fields": 58.95000076293945, "explorer_partner_vision": 0.9505000710487366, "explorer_return": 20.375, "explorer_subspace_norm": 52.64344787597656, "final_distance": 1.7701772451400757, "max_param_abs": 0.8849760293960571, "policy_loss": -0.004696447293270012, "value_loss": 5.442556858062744}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:30.028742+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989430.0287538, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1643355898983363, "elapsed_seconds": 4191.608123634011, "env_steps_per_second": 10497.164501593054, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9526382099168205, "eta_seconds": 3429.4975557005537, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0011881551559200787, "chaser_grad_norm": 0.3450179882347584, "chaser_new_fields": 46.54999923706055, "chaser_partner_vision": 0.9472500681877136, "chaser_return": 25.42500114440918, "chaser_subspace_norm": 47.71327209472656, "collisions_per_episode": 24.600000381469727, "entropy": 0.18037360161542892, "explorer_grad_norm": 0.7127317637205124, "explorer_new_fields": 59.5, "explorer_partner_vision": 0.9472500681877136, "explorer_return": 19.149999618530273, "explorer_subspace_norm": 57.942928314208984, "final_distance": 1.5219706296920776, "max_param_abs": 0.9112007021903992, "policy_loss": -0.005993583955802961, "value_loss": 5.678844690322876}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:53:41.061508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989621.0615225, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2164962188772106, "elapsed_seconds": 4379.3863879579585, "env_steps_per_second": 10503.75461879469, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8991493791701123, "eta_seconds": 3236.9377650124043, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0008681297735337878, "chaser_grad_norm": 0.2752043716609478, "chaser_new_fields": 48.375, "chaser_partner_vision": 0.9337500333786011, "chaser_return": 21.654998779296875, "chaser_subspace_norm": 48.097572326660156, "collisions_per_episode": 20.80000114440918, "entropy": 0.15905753523111343, "explorer_grad_norm": 0.8473901450634003, "explorer_new_fields": 57.875, "explorer_partner_vision": 0.9337500333786011, "explorer_return": 19.475000381469727, "explorer_subspace_norm": 56.994407653808594, "final_distance": 1.7757275104522705, "max_param_abs": 0.9559577703475952, "policy_loss": -0.003888724705198854, "value_loss": 5.280937671661377}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:56:48.839772+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989808.8397827, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2697422567469443, "elapsed_seconds": 4571.0721242889995, "env_steps_per_second": 10500.818778366156, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8464948378312962, "eta_seconds": 3047.381416192666, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0002847606103668632, "chaser_grad_norm": 0.42155007645487785, "chaser_new_fields": 50.150001525878906, "chaser_partner_vision": 0.9100000262260437, "chaser_return": 24.625, "chaser_subspace_norm": 46.80323028564453, "collisions_per_episode": 23.350000381469727, "entropy": 0.1463482566177845, "explorer_grad_norm": 0.43873974680900574, "explorer_new_fields": 60.775001525878906, "explorer_partner_vision": 0.9100000262260437, "explorer_return": 20.0, "explorer_subspace_norm": 63.05351257324219, "final_distance": 1.6251953840255737, "max_param_abs": 0.9797149896621704, "policy_loss": -0.003416150481438507, "value_loss": 5.122330665588379}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:00:00.525509+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990000.5255225, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.321762626705814, "elapsed_seconds": 4758.345456140931, "env_steps_per_second": 10507.854139819125, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7930575760234885, "eta_seconds": 2855.0072736845586, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0001744141228282592, "chaser_grad_norm": 0.4732225835323334, "chaser_new_fields": 50.07500076293945, "chaser_partner_vision": 0.9432500600814819, "chaser_return": 21.247501373291016, "chaser_subspace_norm": 41.869441986083984, "collisions_per_episode": 20.225000381469727, "entropy": 0.14595071226358414, "explorer_grad_norm": 1.348110392689705, "explorer_new_fields": 61.07500076293945, "explorer_partner_vision": 0.9432500600814819, "explorer_return": 23.212499618530273, "explorer_subspace_norm": 71.13052368164062, "final_distance": 1.9023007154464722, "max_param_abs": 0.9905945658683777, "policy_loss": -0.0029910971834326006, "value_loss": 4.9653085470199585}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:03:07.798839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990187.798851, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3737017279858184, "elapsed_seconds": 4945.326220748946, "env_steps_per_second": 10514.978725129451, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7396855458385175, "eta_seconds": 2662.867965018663, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -0.0003033493497337503, "chaser_grad_norm": 0.24047018587589264, "chaser_new_fields": 46.75, "chaser_partner_vision": 0.95250004529953, "chaser_return": 19.652502059936523, "chaser_subspace_norm": 41.72337341308594, "collisions_per_episode": 19.225000381469727, "entropy": 0.15962594002485275, "explorer_grad_norm": 0.16596699133515358, "explorer_new_fields": 50.375, "explorer_partner_vision": 0.95250004529953, "explorer_return": 10.25, "explorer_subspace_norm": 67.01416015625, "final_distance": 1.9316692352294922, "max_param_abs": 0.9943450093269348, "policy_loss": -0.0029906667903993878, "value_loss": 5.365984916687012}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:06:14.779607+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990374.7796214, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4270279376841628, "elapsed_seconds": 5137.300575662986, "env_steps_per_second": 10511.356928542402, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.687087525551634, "eta_seconds": 2473.5150919858825, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0006751392728235883, "chaser_grad_norm": 1.0399435311555862, "chaser_new_fields": 47.67499923706055, "chaser_partner_vision": 0.9415000677108765, "chaser_return": 20.707502365112305, "chaser_subspace_norm": 41.41705322265625, "collisions_per_episode": 20.024999618530273, "entropy": 0.1841299682855606, "explorer_grad_norm": 0.5010040774941444, "explorer_new_fields": 60.125, "explorer_partner_vision": 0.9415000677108765, "explorer_return": 23.762500762939453, "explorer_subspace_norm": 61.62998962402344, "final_distance": 1.677188515663147, "max_param_abs": 0.9638085961341858, "policy_loss": -0.004511416253169642, "value_loss": 5.468553185462952}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:09:26.753960+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990566.753973, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.480260552981668, "elapsed_seconds": 5328.937990734004, "env_steps_per_second": 10508.66046806572, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6343973798492862, "eta_seconds": 2283.8305674574303, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0012872306817008905, "chaser_grad_norm": 0.59980184212327, "chaser_new_fields": 44.75, "chaser_partner_vision": 0.9417500495910645, "chaser_return": 20.075002670288086, "chaser_subspace_norm": 46.37214279174805, "collisions_per_episode": 19.80000114440918, "entropy": 0.15814654901623726, "explorer_grad_norm": 0.2958461865782738, "explorer_new_fields": 60.400001525878906, "explorer_partner_vision": 0.9417500495910645, "explorer_return": 24.212499618530273, "explorer_subspace_norm": 78.11469268798828, "final_distance": 2.138514280319214, "max_param_abs": 0.951306939125061, "policy_loss": -0.00453819336742356, "value_loss": 5.046323657035828}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:12:38.391375+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990758.3913903, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.533622872616947, "elapsed_seconds": 5521.042341421009, "env_steps_per_second": 10505.262668402562, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5817190206478076, "eta_seconds": 2094.188474332107, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0002401197146557621, "chaser_grad_norm": 0.3101242817938328, "chaser_new_fields": 46.125, "chaser_partner_vision": 0.9105000495910645, "chaser_return": 16.755002975463867, "chaser_subspace_norm": 47.842201232910156, "collisions_per_episode": 16.700000762939453, "entropy": 0.1769869141280651, "explorer_grad_norm": 1.32860466837883, "explorer_new_fields": 62.42499923706055, "explorer_partner_vision": 0.9105000495910645, "explorer_return": 28.012500762939453, "explorer_subspace_norm": 65.94043731689453, "final_distance": 1.9316357374191284, "max_param_abs": 1.0022988319396973, "policy_loss": -0.004193592068741481, "value_loss": 5.334564447402954}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:15:50.495725+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990950.4957376, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5875751820513848, "elapsed_seconds": 5715.270655384986, "env_steps_per_second": 10498.190482626995, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5291917273504616, "eta_seconds": 1905.0902184616618, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00040432924864042263, "chaser_grad_norm": 0.19178259372711182, "chaser_new_fields": 51.525001525878906, "chaser_partner_vision": 0.9410000443458557, "chaser_return": 21.935001373291016, "chaser_subspace_norm": 53.29346466064453, "collisions_per_episode": 20.399999618530273, "entropy": 0.13001544028520584, "explorer_grad_norm": 0.43882326781749725, "explorer_new_fields": 61.650001525878906, "explorer_partner_vision": 0.9410000443458557, "explorer_return": 24.587499618530273, "explorer_subspace_norm": 65.88394165039062, "final_distance": 1.6484335660934448, "max_param_abs": 1.0844533443450928, "policy_loss": -0.0035379768526198063, "value_loss": 5.086824297904968}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:19:04.724042+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991144.7240589, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6399221534377688, "elapsed_seconds": 5903.719752375968, "env_steps_per_second": 10501.853509399381, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4761064316432232, "eta_seconds": 1713.9831539156035, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0006722480572607381, "chaser_grad_norm": 0.25304848700761795, "chaser_new_fields": 47.625, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 24.58500099182129, "chaser_subspace_norm": 52.889305114746094, "collisions_per_episode": 23.649999618530273, "entropy": 0.1582392454147339, "explorer_grad_norm": 0.4138009175658226, "explorer_new_fields": 58.150001525878906, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 17.48750114440918, "explorer_subspace_norm": 62.70935821533203, "final_distance": 1.642175316810608, "max_param_abs": 1.0861430168151855, "policy_loss": -0.0038710982063787647, "value_loss": 4.987758755683899}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:22:13.173136+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991333.1731465, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6922923201394344, "elapsed_seconds": 6092.252352501964, "env_steps_per_second": 10505.14592911052, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4230730800348586, "eta_seconds": 1523.063088125491, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.000937797998819434, "chaser_grad_norm": 0.37043919414281845, "chaser_new_fields": 49.47500228881836, "chaser_partner_vision": 0.9355000257492065, "chaser_return": 21.502500534057617, "chaser_subspace_norm": 49.84303665161133, "collisions_per_episode": 20.375, "entropy": 0.16150260716676712, "explorer_grad_norm": 0.47699956223368645, "explorer_new_fields": 60.650001525878906, "explorer_partner_vision": 0.9355000257492065, "explorer_return": 22.200000762939453, "explorer_subspace_norm": 68.17028045654297, "final_distance": 1.6058281660079956, "max_param_abs": 1.0847492218017578, "policy_loss": -0.004623445265927728, "value_loss": 4.7969242334365845}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:25:21.705739+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991521.7057545, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.744404412825275, "elapsed_seconds": 6279.85588617099, "env_steps_per_second": 10509.795319561405, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3700251784780886, "eta_seconds": 1332.090642521119, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0007537112274446089, "chaser_grad_norm": 0.36552849784493446, "chaser_new_fields": 48.875, "chaser_partner_vision": 0.9187500476837158, "chaser_return": 19.232501983642578, "chaser_subspace_norm": 48.18573760986328, "collisions_per_episode": 18.475000381469727, "entropy": 0.18662511557340622, "explorer_grad_norm": 0.48355720192193985, "explorer_new_fields": 61.42499923706055, "explorer_partner_vision": 0.9187500476837158, "explorer_return": 24.98750114440918, "explorer_subspace_norm": 67.38263702392578, "final_distance": 1.6839847564697266, "max_param_abs": 1.1017718315124512, "policy_loss": -0.005853890075347756, "value_loss": 5.467620611190796}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:28:29.309270+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991709.309282, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.7982314320430548, "elapsed_seconds": 6473.633155354997, "env_steps_per_second": 10504.14788236036, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31733495859583316, "eta_seconds": 1142.4058509449994, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00035627437514307303, "chaser_grad_norm": 0.5891587361693382, "chaser_new_fields": 54.07500076293945, "chaser_partner_vision": 0.9262500405311584, "chaser_return": 20.752500534057617, "chaser_subspace_norm": 49.276023864746094, "collisions_per_episode": 19.125, "entropy": 0.1021770928055048, "explorer_grad_norm": 0.4816933944821358, "explorer_new_fields": 61.47500228881836, "explorer_partner_vision": 0.9262500405311584, "explorer_return": 27.887500762939453, "explorer_subspace_norm": 86.1246109008789, "final_distance": 1.8008174896240234, "max_param_abs": 1.1049888134002686, "policy_loss": -0.0016414970380722593, "value_loss": 5.0510454177856445}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:31:43.086540+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991903.0865548, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8507724348986003, "elapsed_seconds": 6662.780765634961, "env_steps_per_second": 10506.123863634139, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26439606212837147, "eta_seconds": 951.8258236621373, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0016457572833200595, "chaser_grad_norm": 0.5161425918340683, "chaser_new_fields": 48.54999923706055, "chaser_partner_vision": 0.9352500438690186, "chaser_return": 20.705001831054688, "chaser_subspace_norm": 51.71734619140625, "collisions_per_episode": 19.75, "entropy": 0.1614171378314495, "explorer_grad_norm": 0.1964201256632805, "explorer_new_fields": 58.0, "explorer_partner_vision": 0.9352500438690186, "explorer_return": 20.975000381469727, "explorer_subspace_norm": 82.28788757324219, "final_distance": 1.5810810327529907, "max_param_abs": 1.1334906816482544, "policy_loss": -0.0055989406332415115, "value_loss": 5.6854753494262695}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:34:52.234149+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992092.2341604, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9033792166402823, "elapsed_seconds": 6852.165179905016, "env_steps_per_second": 10507.627605235, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.211486579626698, "eta_seconds": 761.3516866561129, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": -0.00022323531053114465, "chaser_grad_norm": 0.43752819299697876, "chaser_new_fields": 48.025001525878906, "chaser_partner_vision": 0.9425000548362732, "chaser_return": 19.340002059936523, "chaser_subspace_norm": 53.68885040283203, "collisions_per_episode": 18.5, "entropy": 0.12374857999384403, "explorer_grad_norm": 0.5627283528447151, "explorer_new_fields": 58.25, "explorer_partner_vision": 0.9425000548362732, "explorer_return": 21.075000762939453, "explorer_subspace_norm": 96.07400512695312, "final_distance": 1.5895527601242065, "max_param_abs": 1.1357388496398926, "policy_loss": -0.003098882641644707, "value_loss": 4.997665286064148}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:38:01.618566+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992281.618582, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.955654452318882, "elapsed_seconds": 7040.356028347975, "env_steps_per_second": 10510.832080371958, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1585665772150445, "eta_seconds": 570.8396779741602, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0003212530290990223, "chaser_grad_norm": 1.2954601049423218, "chaser_new_fields": 42.85000228881836, "chaser_partner_vision": 0.9320000410079956, "chaser_return": 22.192501068115234, "chaser_subspace_norm": 50.33931350708008, "collisions_per_episode": 21.92500114440918, "entropy": 0.17458152398467064, "explorer_grad_norm": 0.48033298552036285, "explorer_new_fields": 50.0, "explorer_partner_vision": 0.9320000410079956, "explorer_return": 9.425000190734863, "explorer_subspace_norm": 100.84282684326172, "final_distance": 1.6751664876937866, "max_param_abs": 1.1677584648132324, "policy_loss": -0.0036471000732873904, "value_loss": 5.543139100074768}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:41:09.809412+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992469.809424, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.007479782194423, "elapsed_seconds": 7226.927215899923, "env_steps_per_second": 10516.226015503908, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10565683064181174, "eta_seconds": 380.36459031052226, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005578143868472107, "chaser_grad_norm": 0.576453622430563, "chaser_new_fields": 45.525001525878906, "chaser_partner_vision": 0.9452500343322754, "chaser_return": 18.052501678466797, "chaser_subspace_norm": 51.30323028564453, "collisions_per_episode": 18.024999618530273, "entropy": 0.1504194512963295, "explorer_grad_norm": 0.3156035915017128, "explorer_new_fields": 55.5, "explorer_partner_vision": 0.9452500343322754, "explorer_return": 18.012500762939453, "explorer_subspace_norm": 88.64036560058594, "final_distance": 1.5152206420898438, "max_param_abs": 1.1654928922653198, "policy_loss": -0.003493720475583828, "value_loss": 5.344757556915283}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:44:16.380600+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992656.3806138, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.061037331766921, "elapsed_seconds": 7419.734394360916, "env_steps_per_second": 10512.505684742691, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.052847111070946695, "eta_seconds": 190.2495998554081, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005213677731781274, "chaser_grad_norm": 0.8448120132088661, "chaser_new_fields": 49.35000228881836, "chaser_partner_vision": 0.9250000715255737, "chaser_return": 19.639999389648438, "chaser_subspace_norm": 51.463768005371094, "collisions_per_episode": 18.75, "entropy": 0.17568372562527657, "explorer_grad_norm": 0.33188436925411224, "explorer_new_fields": 58.04999923706055, "explorer_partner_vision": 0.9250000715255737, "explorer_return": 21.662500381469727, "explorer_subspace_norm": 106.46646118164062, "final_distance": 1.9493179321289062, "max_param_abs": 1.2215321063995361, "policy_loss": -0.004636985326186505, "value_loss": 5.504049062728882}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:47:29.187779+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992849.1877923, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1074996488380324, "elapsed_seconds": 7586.998735816916, "env_steps_per_second": 10544.353938314733, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:50:16.452119+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993016.4521298, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.10750246579028, "elapsed_seconds": 7587.008876845008, "env_steps_per_second": 10544.33984440879, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.00013653355120357702, "chaser_grad_norm": 0.9264960587024689, "chaser_new_fields": 46.92499923706055, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 24.285001754760742, "chaser_subspace_norm": 67.47857666015625, "collisions_per_episode": 23.200000762939453, "entropy": 0.1407805271446705, "explorer_grad_norm": 0.5267750471830368, "explorer_new_fields": 55.900001525878906, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 17.149999618530273, "explorer_subspace_norm": 89.71617889404297, "final_distance": 1.689988374710083, "max_param_abs": 1.2198518514633179, "policy_loss": -0.003363337086513596, "value_loss": 4.716352343559265}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:50:16.462261+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993016.4622712, "update": 20000} diff --git a/social/seed_0007/attempt_01/status.json b/social/seed_0007/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..cd8be6ce598de7718d217ddb8b4d250005d61e5f --- /dev/null +++ b/social/seed_0007/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.10750246579028, + "elapsed_seconds": 7587.008876845008, + "env_steps_per_second": 10544.33984440879, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.00013653355120357702, + "chaser_grad_norm": 0.9264960587024689, + "chaser_new_fields": 46.92499923706055, + "chaser_partner_vision": 0.9470000267028809, + "chaser_return": 24.285001754760742, + "chaser_subspace_norm": 67.47857666015625, + "collisions_per_episode": 23.200000762939453, + "entropy": 0.1407805271446705, + "explorer_grad_norm": 0.5267750471830368, + "explorer_new_fields": 55.900001525878906, + "explorer_partner_vision": 0.9470000267028809, + "explorer_return": 17.149999618530273, + "explorer_subspace_norm": 89.71617889404297, + "final_distance": 1.689988374710083, + "max_param_abs": 1.2198518514633179, + "policy_loss": -0.003363337086513596, + "value_loss": 4.716352343559265 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:16.462261+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782993016.4622712, + "update": 20000 +} diff --git a/social/seed_0008/attempt_01/checkpoints/latest.safetensors b/social/seed_0008/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b7108d432beeb39f649955e51f95453ee88791ed --- /dev/null +++ b/social/seed_0008/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:39e2f67475bf8a4eca5f24de315185f9f759d1f65ac80f68a5644a58f16272e1 +size 11460392 diff --git a/social/seed_0008/attempt_01/config.json b/social/seed_0008/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..60227947434253e2f0d0d64a4ecabf81ad27433c --- /dev/null +++ b/social/seed_0008/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0008.attempt_01", + "seed": 8, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0008/attempt_01/metrics.jsonl b/social/seed_0008/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..2e60f2d48a92fbb4c9bc7df996a96c656b1c9c44 --- /dev/null +++ b/social/seed_0008/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005330539427814074, "elapsed_seconds": 1.9189941940130666, "env_steps_per_second": 2084.425274698233, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.660545801685366, "eta_seconds": 38377.96488606732, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 5.5856855824820695e-06, "chaser_grad_norm": 0.242379542440176, "chaser_new_fields": 33.45000076293945, "chaser_partner_vision": 0.31300002336502075, "chaser_return": -2.015000343322754, "chaser_subspace_norm": 0.004576629027724266, "collisions_per_episode": 1.2000000476837158, "entropy": 2.772408902645111, "explorer_grad_norm": 0.14950605854392052, "explorer_new_fields": 33.150001525878906, "explorer_partner_vision": 0.31300002336502075, "explorer_return": -1.0625, "explorer_subspace_norm": 0.0026072447653859854, "final_distance": 5.316863536834717, "max_param_abs": 0.15428009629249573, "policy_loss": -0.0008525772547587351, "value_loss": 3.141881048679352}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:44:01.671603+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985441.6716173, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05352810830999968, "elapsed_seconds": 192.70118991599884, "env_steps_per_second": 10378.763104015228, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0875962240899875, "eta_seconds": 7515.346406723955, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0004958402200687106, "chaser_grad_norm": 0.26819054782390594, "chaser_new_fields": 59.29999923706055, "chaser_partner_vision": 0.5502500534057617, "chaser_return": 9.097500801086426, "chaser_subspace_norm": 14.87891674041748, "collisions_per_episode": 6.825000286102295, "entropy": 0.6709342449903488, "explorer_grad_norm": 0.27240126579999924, "explorer_new_fields": 63.10000228881836, "explorer_partner_vision": 0.5502500534057617, "explorer_return": 39.32500076293945, "explorer_subspace_norm": 7.7286834716796875, "final_distance": 4.286420822143555, "max_param_abs": 0.3212723731994629, "policy_loss": -0.0058134156889195765, "value_loss": 5.484553575515747}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:47:12.453797+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985632.45381, "update": 500} +{"device": "cuda", "elapsed_hours": 0.1057016417214052, "elapsed_seconds": 380.5259101970587, "env_steps_per_second": 10511.767774048723, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0083311927066987, "eta_seconds": 7229.992293744115, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0006020753804186763, "chaser_grad_norm": 0.28860072791576385, "chaser_new_fields": 53.32500076293945, "chaser_partner_vision": 0.7292500138282776, "chaser_return": 12.987500190734863, "chaser_subspace_norm": 23.81314468383789, "collisions_per_episode": 11.625, "entropy": 0.5304818004369736, "explorer_grad_norm": 0.7088412344455719, "explorer_new_fields": 62.72500228881836, "explorer_partner_vision": 0.7292500138282776, "explorer_return": 35.20000076293945, "explorer_subspace_norm": 9.647390365600586, "final_distance": 2.637505054473877, "max_param_abs": 0.37493273615837097, "policy_loss": -0.004307829149765574, "value_loss": 5.647135019302368}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:50:20.278517+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985820.278531, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.15768789607863357, "elapsed_seconds": 567.6764258830808, "env_steps_per_second": 10569.401381546477, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9448173849698138, "eta_seconds": 7001.34258589133, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0002908518923216974, "chaser_grad_norm": 0.34635037928819656, "chaser_new_fields": 47.275001525878906, "chaser_partner_vision": 0.8007500171661377, "chaser_return": 12.210000038146973, "chaser_subspace_norm": 29.634794235229492, "collisions_per_episode": 12.0, "entropy": 0.41872137784957886, "explorer_grad_norm": 0.4481799229979515, "explorer_new_fields": 59.525001525878906, "explorer_partner_vision": 0.8007500171661377, "explorer_return": 29.575000762939453, "explorer_subspace_norm": 14.415115356445312, "final_distance": 1.9479912519454956, "max_param_abs": 0.4019787609577179, "policy_loss": -0.006600278138648719, "value_loss": 5.450750708580017}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:27.429032+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986007.4290433, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.2119841591297235, "elapsed_seconds": 763.1429728670046, "env_steps_per_second": 10482.963591927337, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9078574321675115, "eta_seconds": 6868.286755803041, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 5.357617424417427e-05, "chaser_grad_norm": 0.45206793397665024, "chaser_new_fields": 51.775001525878906, "chaser_partner_vision": 0.8682500123977661, "chaser_return": 11.75, "chaser_subspace_norm": 35.12519454956055, "collisions_per_episode": 10.90000057220459, "entropy": 0.293507881462574, "explorer_grad_norm": 0.5496172904968262, "explorer_new_fields": 62.97500228881836, "explorer_partner_vision": 0.8682500123977661, "explorer_return": 35.07500076293945, "explorer_subspace_norm": 20.1675968170166, "final_distance": 2.370321035385132, "max_param_abs": 0.46341991424560547, "policy_loss": -0.006259505898697171, "value_loss": 5.021396279335022}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:42.895583+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986202.895596, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.2662556969341757, "elapsed_seconds": 958.5205089630326, "env_steps_per_second": 10432.744950672382, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.86378987853923, "eta_seconds": 6709.6435627412275, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.001218862790420161, "chaser_grad_norm": 0.32507631182670593, "chaser_new_fields": 50.275001525878906, "chaser_partner_vision": 0.9025000333786011, "chaser_return": 13.51500129699707, "chaser_subspace_norm": 38.068321228027344, "collisions_per_episode": 12.75, "entropy": 0.2980204150080681, "explorer_grad_norm": 0.3272622749209404, "explorer_new_fields": 62.0, "explorer_partner_vision": 0.9025000333786011, "explorer_return": 31.9375, "explorer_subspace_norm": 21.765785217285156, "final_distance": 1.8083946704864502, "max_param_abs": 0.46946731209754944, "policy_loss": -0.007486004411248359, "value_loss": 5.3979973793029785}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:58.273116+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986398.2731268, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.3195249417549995, "elapsed_seconds": 1150.2897903179983, "env_steps_per_second": 10432.153793769301, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.810641336611664, "eta_seconds": 6518.3088118019905, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0010391328742465067, "chaser_grad_norm": 0.2863578498363495, "chaser_new_fields": 49.650001525878906, "chaser_partner_vision": 0.8917500376701355, "chaser_return": 14.720001220703125, "chaser_subspace_norm": 37.218955993652344, "collisions_per_episode": 13.90000057220459, "entropy": 0.23999937251210213, "explorer_grad_norm": 0.2654268443584442, "explorer_new_fields": 61.900001525878906, "explorer_partner_vision": 0.8917500376701355, "explorer_return": 30.649999618530273, "explorer_subspace_norm": 28.39801025390625, "final_distance": 2.0644500255584717, "max_param_abs": 0.4977448284626007, "policy_loss": -0.006361737969797332, "value_loss": 5.417000889778137}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:03:10.042397+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986590.042409, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.3720014799230719, "elapsed_seconds": 1339.2053277230589, "env_steps_per_second": 10453.960800621257, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7537212624944818, "eta_seconds": 6313.396544980134, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0017684021518196147, "chaser_grad_norm": 0.3688257187604904, "chaser_new_fields": 47.82500076293945, "chaser_partner_vision": 0.8692500591278076, "chaser_return": 15.097500801086426, "chaser_subspace_norm": 38.758941650390625, "collisions_per_episode": 14.675000190734863, "entropy": 0.21027911081910133, "explorer_grad_norm": 0.3385261371731758, "explorer_new_fields": 63.025001525878906, "explorer_partner_vision": 0.8692500591278076, "explorer_return": 31.762500762939453, "explorer_subspace_norm": 28.360013961791992, "final_distance": 2.216040849685669, "max_param_abs": 0.5736011266708374, "policy_loss": -0.006215131744742308, "value_loss": 5.581642031669617}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:06:18.957935+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986778.957948, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.42396954483805327, "elapsed_seconds": 1526.2903614169918, "env_steps_per_second": 10482.933263855359, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.695878179352213, "eta_seconds": 6105.161445667967, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0008406550980009087, "chaser_grad_norm": 0.2781466692686081, "chaser_new_fields": 49.47500228881836, "chaser_partner_vision": 0.8982500433921814, "chaser_return": 16.340002059936523, "chaser_subspace_norm": 40.19259262084961, "collisions_per_episode": 15.600000381469727, "entropy": 0.17337730899453163, "explorer_grad_norm": 0.29922056943178177, "explorer_new_fields": 62.400001525878906, "explorer_partner_vision": 0.8982500433921814, "explorer_return": 29.98750114440918, "explorer_subspace_norm": 31.89286231994629, "final_distance": 2.1185872554779053, "max_param_abs": 0.5663326978683472, "policy_loss": -0.00492162701161547, "value_loss": 5.857679009437561}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:09:26.042968+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986966.0429828, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4768680220733384, "elapsed_seconds": 1716.7248794640182, "env_steps_per_second": 10485.08134024353, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.642545409363721, "eta_seconds": 5913.1634737093955, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00023713635495203889, "chaser_grad_norm": 0.23348581418395042, "chaser_new_fields": 47.875, "chaser_partner_vision": 0.8892500400543213, "chaser_return": 15.730000495910645, "chaser_subspace_norm": 38.60639953613281, "collisions_per_episode": 15.15000057220459, "entropy": 0.1809796206653118, "explorer_grad_norm": 0.44747279584407806, "explorer_new_fields": 61.0, "explorer_partner_vision": 0.8892500400543213, "explorer_return": 28.712499618530273, "explorer_subspace_norm": 41.64932632446289, "final_distance": 1.9393523931503296, "max_param_abs": 0.5724173188209534, "policy_loss": -0.004493310282007279, "value_loss": 5.343045592308044}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:36.477485+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987156.4774964, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5302664292708505, "elapsed_seconds": 1908.9591453750618, "env_steps_per_second": 10476.913583224177, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5907992878125512, "eta_seconds": 5726.877436125184, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0007805787517511931, "chaser_grad_norm": 2.245885580778122, "chaser_new_fields": 49.29999923706055, "chaser_partner_vision": 0.8890000581741333, "chaser_return": 17.990001678466797, "chaser_subspace_norm": 38.787147521972656, "collisions_per_episode": 17.0, "entropy": 0.1881377249956131, "explorer_grad_norm": 0.8784344494342804, "explorer_new_fields": 61.70000076293945, "explorer_partner_vision": 0.8890000581741333, "explorer_return": 28.762500762939453, "explorer_subspace_norm": 44.98735427856445, "final_distance": 1.8602821826934814, "max_param_abs": 0.6856481432914734, "policy_loss": -0.00486253396468328, "value_loss": 5.813503265380859}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:48.711753+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987348.7117696, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5840104481547476, "elapsed_seconds": 2102.4376133570913, "env_steps_per_second": 10464.044145819504, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5396639087716073, "eta_seconds": 5542.790071577786, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.000622467230639856, "chaser_grad_norm": 0.24560433626174927, "chaser_new_fields": 50.92499923706055, "chaser_partner_vision": 0.9145000576972961, "chaser_return": 21.22749900817871, "chaser_subspace_norm": 38.26223373413086, "collisions_per_episode": 19.67500114440918, "entropy": 0.15449003130197525, "explorer_grad_norm": 0.5233577340841293, "explorer_new_fields": 65.4749984741211, "explorer_partner_vision": 0.9145000576972961, "explorer_return": 32.03750228881836, "explorer_subspace_norm": 52.60789108276367, "final_distance": 1.8698230981826782, "max_param_abs": 0.6838610768318176, "policy_loss": -0.004536007163110867, "value_loss": 5.34645688533783}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:19:02.190220+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987542.1902308, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6373663621188865, "elapsed_seconds": 2294.5189036279917, "env_steps_per_second": 10459.708988255561, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.487188178277402, "eta_seconds": 5353.877441798647, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.001400804207420947, "chaser_grad_norm": 0.6707736998796463, "chaser_new_fields": 48.625, "chaser_partner_vision": 0.9332500696182251, "chaser_return": 19.537500381469727, "chaser_subspace_norm": 36.69968032836914, "collisions_per_episode": 18.524999618530273, "entropy": 0.2081972062587738, "explorer_grad_norm": 0.3851800858974457, "explorer_new_fields": 58.45000076293945, "explorer_partner_vision": 0.9332500696182251, "explorer_return": 22.48750114440918, "explorer_subspace_norm": 47.67208480834961, "final_distance": 1.7085565328598022, "max_param_abs": 0.6968445181846619, "policy_loss": -0.0060098354546340715, "value_loss": 5.769705057144165}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:22:14.271511+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987734.2715247, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6902971984686317, "elapsed_seconds": 2485.069914487074, "env_steps_per_second": 10462.482302179606, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4336941814348507, "eta_seconds": 5161.299053165462, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0005494766699314368, "chaser_grad_norm": 0.7745308056473732, "chaser_new_fields": 48.04999923706055, "chaser_partner_vision": 0.9177500605583191, "chaser_return": 19.540000915527344, "chaser_subspace_norm": 36.49509048461914, "collisions_per_episode": 19.0, "entropy": 0.1915583796799183, "explorer_grad_norm": 0.2413507215678692, "explorer_new_fields": 60.22500228881836, "explorer_partner_vision": 0.9177500605583191, "explorer_return": 23.5625, "explorer_subspace_norm": 46.380218505859375, "final_distance": 1.920669436454773, "max_param_abs": 0.7108176350593567, "policy_loss": -0.005302781247532806, "value_loss": 5.712437987327576}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:25:24.822520+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987924.8225327, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7436480285825221, "elapsed_seconds": 2677.1329028970795, "env_steps_per_second": 10458.950308256863, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.381060624510398, "eta_seconds": 4971.818248237433, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0015098362873295912, "chaser_grad_norm": 0.4213273972272873, "chaser_new_fields": 48.375, "chaser_partner_vision": 0.893500030040741, "chaser_return": 19.0575008392334, "chaser_subspace_norm": 35.9431266784668, "collisions_per_episode": 18.225000381469727, "entropy": 0.18496929109096527, "explorer_grad_norm": 0.541811503469944, "explorer_new_fields": 59.150001525878906, "explorer_partner_vision": 0.893500030040741, "explorer_return": 23.274999618530273, "explorer_subspace_norm": 52.42327117919922, "final_distance": 1.8067600727081299, "max_param_abs": 0.7267396450042725, "policy_loss": -0.005059673362410422, "value_loss": 5.606943368911743}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:28:36.885510+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988116.8855243, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7978458191858308, "elapsed_seconds": 2872.2449490689905, "env_steps_per_second": 10444.791628835208, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3297430319763843, "eta_seconds": 4787.074915114984, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0010673934837410748, "chaser_grad_norm": 0.2516757510602474, "chaser_new_fields": 49.75, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 17.57000160217285, "chaser_subspace_norm": 37.16876983642578, "collisions_per_episode": 16.700000762939453, "entropy": 0.17442208901047707, "explorer_grad_norm": 0.23168740421533585, "explorer_new_fields": 62.92499923706055, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 30.11250114440918, "explorer_subspace_norm": 52.55442810058594, "final_distance": 1.8432130813598633, "max_param_abs": 0.7932367920875549, "policy_loss": -0.0056478826236006086, "value_loss": 5.406368732452393}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:31:51.997556+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988311.9975698, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8506366340308321, "elapsed_seconds": 3062.2918825109955, "env_steps_per_second": 10449.689718591057, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2759549510462482, "eta_seconds": 4593.437823766493, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0008644843299454585, "chaser_grad_norm": 0.21359294280409813, "chaser_new_fields": 51.650001525878906, "chaser_partner_vision": 0.9380000233650208, "chaser_return": 16.610000610351562, "chaser_subspace_norm": 38.31128692626953, "collisions_per_episode": 15.40000057220459, "entropy": 0.193915206938982, "explorer_grad_norm": 0.48032183945178986, "explorer_new_fields": 63.07500076293945, "explorer_partner_vision": 0.9380000233650208, "explorer_return": 31.625, "explorer_subspace_norm": 54.89335632324219, "final_distance": 1.7696783542633057, "max_param_abs": 0.7892124056816101, "policy_loss": -0.006090335871094865, "value_loss": 5.285418152809143}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:35:02.044490+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988502.0445042, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9028573722008474, "elapsed_seconds": 3250.2865399230504, "env_steps_per_second": 10460.616189489847, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2215129153305584, "eta_seconds": 4397.44649519001, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0006316838725865526, "chaser_grad_norm": 0.42924345657229424, "chaser_new_fields": 49.650001525878906, "chaser_partner_vision": 0.909000039100647, "chaser_return": 15.957501411437988, "chaser_subspace_norm": 39.76167297363281, "collisions_per_episode": 15.225000381469727, "entropy": 0.18809866160154343, "explorer_grad_norm": 0.30972379446029663, "explorer_new_fields": 58.57500076293945, "explorer_partner_vision": 0.909000039100647, "explorer_return": 24.4375, "explorer_subspace_norm": 58.44508743286133, "final_distance": 1.9763659238815308, "max_param_abs": 0.7872291803359985, "policy_loss": -0.00476884692162205, "value_loss": 5.612807989120483}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:38:10.039146+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988690.0391562, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9546127292600188, "elapsed_seconds": 3436.605825336068, "env_steps_per_second": 10475.452184417902, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.166748891317801, "eta_seconds": 4200.296008744083, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0004755821987894038, "chaser_grad_norm": 0.4977376013994217, "chaser_new_fields": 48.07500076293945, "chaser_partner_vision": 0.9465000629425049, "chaser_return": 21.172500610351562, "chaser_subspace_norm": 38.75114059448242, "collisions_per_episode": 20.575000762939453, "entropy": 0.19150183349847794, "explorer_grad_norm": 0.7479203790426254, "explorer_new_fields": 59.10000228881836, "explorer_partner_vision": 0.9465000629425049, "explorer_return": 20.412500381469727, "explorer_subspace_norm": 60.85232925415039, "final_distance": 1.8124767541885376, "max_param_abs": 0.7800331115722656, "policy_loss": -0.006105437634502664, "value_loss": 5.468934059143066}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:41:16.358433+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988876.3584466, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0063082131375105, "elapsed_seconds": 3622.709567295038, "env_steps_per_second": 10489.386271274678, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.112235393467775, "eta_seconds": 4004.0474164839893, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.002136922406211772, "chaser_grad_norm": 0.319851890206337, "chaser_new_fields": 49.150001525878906, "chaser_partner_vision": 0.9385000467300415, "chaser_return": 23.704999923706055, "chaser_subspace_norm": 40.44599151611328, "collisions_per_episode": 22.600000381469727, "entropy": 0.1629793532192707, "explorer_grad_norm": 0.2848823145031929, "explorer_new_fields": 58.150001525878906, "explorer_partner_vision": 0.9385000467300415, "explorer_return": 18.125, "explorer_subspace_norm": 66.87946319580078, "final_distance": 1.595103144645691, "max_param_abs": 0.8141506314277649, "policy_loss": -0.005351223132574745, "value_loss": 5.487300753593445}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:44:22.462174+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989062.4621851, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0590291792875117, "elapsed_seconds": 3812.505045435042, "env_steps_per_second": 10491.789393930003, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0590291792875117, "eta_seconds": 3812.505045435042, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005033882107462473, "chaser_grad_norm": 1.412640556693077, "chaser_new_fields": 46.375, "chaser_partner_vision": 0.9317500591278076, "chaser_return": 20.662500381469727, "chaser_subspace_norm": 43.46104431152344, "collisions_per_episode": 20.17500114440918, "entropy": 0.14028999209403992, "explorer_grad_norm": 0.292730875313282, "explorer_new_fields": 58.20000076293945, "explorer_partner_vision": 0.9317500591278076, "explorer_return": 20.42500114440918, "explorer_subspace_norm": 62.04376220703125, "final_distance": 1.6393886804580688, "max_param_abs": 0.910667896270752, "policy_loss": -0.003526977518106378, "value_loss": 5.325456500053406}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:47:32.257652+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989252.2576663, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1130753425094493, "elapsed_seconds": 4007.0712330340175, "env_steps_per_second": 10481.470769412561, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.007068167032359, "eta_seconds": 3625.445401316492, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -9.05828621049487e-05, "chaser_grad_norm": 0.2709626704454422, "chaser_new_fields": 51.42499923706055, "chaser_partner_vision": 0.9450000524520874, "chaser_return": 21.829999923706055, "chaser_subspace_norm": 40.22971725463867, "collisions_per_episode": 20.399999618530273, "entropy": 0.14416120573878288, "explorer_grad_norm": 0.5016419887542725, "explorer_new_fields": 61.79999923706055, "explorer_partner_vision": 0.9450000524520874, "explorer_return": 25.450000762939453, "explorer_subspace_norm": 68.70757293701172, "final_distance": 1.4291974306106567, "max_param_abs": 0.9349232316017151, "policy_loss": -0.0033605736792088337, "value_loss": 5.362581729888916}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:46.823839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989446.8238509, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1667576488486116, "elapsed_seconds": 4200.327535855002, "env_steps_per_second": 10475.373557039413, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9546198945125003, "eta_seconds": 3436.631620245001, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0013249207809258312, "chaser_grad_norm": 0.2508114315569401, "chaser_new_fields": 49.650001525878906, "chaser_partner_vision": 0.9362500309944153, "chaser_return": 19.8700008392334, "chaser_subspace_norm": 42.48298645019531, "collisions_per_episode": 18.850000381469727, "entropy": 0.14305970072746277, "explorer_grad_norm": 0.407844141125679, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.9362500309944153, "explorer_return": 24.012500762939453, "explorer_subspace_norm": 66.22245025634766, "final_distance": 1.7232249975204468, "max_param_abs": 0.947340190410614, "policy_loss": -0.004788745637053893, "value_loss": 5.403303503990173}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:54:00.080143+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989640.080158, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.220831832841666, "elapsed_seconds": 4394.994598229998, "env_steps_per_second": 10466.451999400782, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9023539634047097, "eta_seconds": 3248.474268256955, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0005571391081997668, "chaser_grad_norm": 0.8118366822600365, "chaser_new_fields": 51.45000076293945, "chaser_partner_vision": 0.9440000653266907, "chaser_return": 17.877500534057617, "chaser_subspace_norm": 42.78097152709961, "collisions_per_episode": 16.524999618530273, "entropy": 0.15614653751254082, "explorer_grad_norm": 0.5780624225735664, "explorer_new_fields": 58.29999923706055, "explorer_partner_vision": 0.9440000653266907, "explorer_return": 22.024999618530273, "explorer_subspace_norm": 67.62272644042969, "final_distance": 1.7210887670516968, "max_param_abs": 0.9374839067459106, "policy_loss": -0.0045425964538026165, "value_loss": 5.244042277336121}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:57:14.747206+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989834.7472203, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2739936065361124, "elapsed_seconds": 4586.3769835300045, "env_steps_per_second": 10465.777273078795, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.849329071024075, "eta_seconds": 3057.58465568667, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0011350465408012655, "chaser_grad_norm": 0.24765625968575478, "chaser_new_fields": 50.10000228881836, "chaser_partner_vision": 0.9507500529289246, "chaser_return": 20.250001907348633, "chaser_subspace_norm": 43.66331100463867, "collisions_per_episode": 19.05000114440918, "entropy": 0.1565168760716915, "explorer_grad_norm": 0.23353637382388115, "explorer_new_fields": 58.525001525878906, "explorer_partner_vision": 0.9507500529289246, "explorer_return": 22.5625, "explorer_subspace_norm": 72.94817352294922, "final_distance": 1.9890211820602417, "max_param_abs": 0.9441203474998474, "policy_loss": -0.0052035811301323065, "value_loss": 5.343036890029907}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:00:26.129592+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990026.1296058, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3271771454925247, "elapsed_seconds": 4777.837723773089, "env_steps_per_second": 10464.984976617137, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7963062872955149, "eta_seconds": 2866.7026342638537, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005820272016106021, "chaser_grad_norm": 0.21127810329198837, "chaser_new_fields": 47.17499923706055, "chaser_partner_vision": 0.8917500376701355, "chaser_return": 18.677499771118164, "chaser_subspace_norm": 45.54037094116211, "collisions_per_episode": 18.17500114440918, "entropy": 0.16563476249575615, "explorer_grad_norm": 0.496120922267437, "explorer_new_fields": 62.04999923706055, "explorer_partner_vision": 0.8917500376701355, "explorer_return": 26.48750114440918, "explorer_subspace_norm": 82.74331665039062, "final_distance": 1.8864792585372925, "max_param_abs": 0.9441203474998474, "policy_loss": -0.006206639908254985, "value_loss": 5.339436769485474}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:03:37.590329+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990217.5903409, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3813396838830603, "elapsed_seconds": 4972.822861979017, "env_steps_per_second": 10456.837382561771, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7437982913216479, "eta_seconds": 2677.6738487579323, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.000835279536287814, "chaser_grad_norm": 0.3727348893880844, "chaser_new_fields": 52.04999923706055, "chaser_partner_vision": 0.9377500414848328, "chaser_return": 19.170001983642578, "chaser_subspace_norm": 47.54634475708008, "collisions_per_episode": 17.649999618530273, "entropy": 0.1324244812130928, "explorer_grad_norm": 0.39158639311790466, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9377500414848328, "explorer_return": 25.325000762939453, "explorer_subspace_norm": 73.01494598388672, "final_distance": 1.7065070867538452, "max_param_abs": 0.9640337824821472, "policy_loss": -0.004207860117780626, "value_loss": 5.004323840141296}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:06:52.575469+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990412.5754836, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4347761041827876, "elapsed_seconds": 5165.193975058035, "env_steps_per_second": 10454.592849902267, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6908181242361571, "eta_seconds": 2486.9452472501653, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0008104897331211447, "chaser_grad_norm": 0.28602490201592445, "chaser_new_fields": 47.25, "chaser_partner_vision": 0.9440000653266907, "chaser_return": 18.895002365112305, "chaser_subspace_norm": 50.188743591308594, "collisions_per_episode": 18.5, "entropy": 0.11876515485346317, "explorer_grad_norm": 0.2794200927019119, "explorer_new_fields": 55.92499923706055, "explorer_partner_vision": 0.9440000653266907, "explorer_return": 17.774999618530273, "explorer_subspace_norm": 82.32061004638672, "final_distance": 1.9366447925567627, "max_param_abs": 1.000549554824829, "policy_loss": -0.0026482388088489017, "value_loss": 5.404518485069275}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:10:04.946582+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990604.9465957, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4875908566152793, "elapsed_seconds": 5355.327083815006, "env_steps_per_second": 10456.877632973064, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6375389385494055, "eta_seconds": 2295.1401787778595, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0006175738644473849, "chaser_grad_norm": 1.2750466912984848, "chaser_new_fields": 51.125, "chaser_partner_vision": 0.9260000586509705, "chaser_return": 19.26500129699707, "chaser_subspace_norm": 45.05524444580078, "collisions_per_episode": 17.950000762939453, "entropy": 0.15509118139743805, "explorer_grad_norm": 0.9141453802585602, "explorer_new_fields": 57.275001525878906, "explorer_partner_vision": 0.9260000586509705, "explorer_return": 19.8125, "explorer_subspace_norm": 79.29683685302734, "final_distance": 1.8251659870147705, "max_param_abs": 1.000549554824829, "policy_loss": -0.00305984698654882, "value_loss": 5.689659476280212}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:13:15.079690+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990795.079704, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5399832915753036, "elapsed_seconds": 5543.939849671093, "env_steps_per_second": 10461.873969184746, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5841315933561496, "eta_seconds": 2102.8737360821387, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 2.2348410963074494e-05, "chaser_grad_norm": 0.5067558735609055, "chaser_new_fields": 47.525001525878906, "chaser_partner_vision": 0.9115000367164612, "chaser_return": 18.60500144958496, "chaser_subspace_norm": 46.99578094482422, "collisions_per_episode": 17.950000762939453, "entropy": 0.11412940919399261, "explorer_grad_norm": 0.8045390397310257, "explorer_new_fields": 59.47500228881836, "explorer_partner_vision": 0.9115000367164612, "explorer_return": 23.600000381469727, "explorer_subspace_norm": 79.84294128417969, "final_distance": 1.8500686883926392, "max_param_abs": 1.000549554824829, "policy_loss": -0.0024251421885601587, "value_loss": 4.619950890541077}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:16:23.692456+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990983.6924667, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5918987834261173, "elapsed_seconds": 5730.835620334023, "env_steps_per_second": 10469.67736905755, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5306329278087057, "eta_seconds": 1910.2785401113408, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.013931899792184788, "chaser_grad_norm": 0.47859377786517143, "chaser_new_fields": 46.95000076293945, "chaser_partner_vision": 0.9035000205039978, "chaser_return": 28.75, "chaser_subspace_norm": 53.56529235839844, "collisions_per_episode": 27.30000114440918, "entropy": 0.11137841828167439, "explorer_grad_norm": 0.458875373005867, "explorer_new_fields": 57.875, "explorer_partner_vision": 0.9035000205039978, "explorer_return": 18.36250114440918, "explorer_subspace_norm": 86.54653930664062, "final_distance": 1.604865312576294, "max_param_abs": 1.008047342300415, "policy_loss": 0.002280008649918841, "value_loss": 5.836463093757629}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:19:30.588227+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991170.5882413, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6436084777602809, "elapsed_seconds": 5916.990519937011, "env_steps_per_second": 10478.299701697006, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47717665483363, "eta_seconds": 1717.8359574010678, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": -0.0002866952166469261, "chaser_grad_norm": 0.38459373638033867, "chaser_new_fields": 50.25, "chaser_partner_vision": 0.9017500281333923, "chaser_return": 18.625001907348633, "chaser_subspace_norm": 49.24122619628906, "collisions_per_episode": 17.55000114440918, "entropy": 0.1303776130080223, "explorer_grad_norm": 0.2920645996928215, "explorer_new_fields": 62.04999923706055, "explorer_partner_vision": 0.9017500281333923, "explorer_return": 27.774999618530273, "explorer_subspace_norm": 84.94609832763672, "final_distance": 1.7066048383712769, "max_param_abs": 1.0050793886184692, "policy_loss": -0.0028552790340032885, "value_loss": 4.693346619606018}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:22:36.743126+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991356.7431376, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6954080422830562, "elapsed_seconds": 6103.468952219002, "env_steps_per_second": 10485.840183840355, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.42385201057076405, "eta_seconds": 1525.8672380547505, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00036556952581090485, "chaser_grad_norm": 0.30156490951776505, "chaser_new_fields": 51.72500228881836, "chaser_partner_vision": 0.9137500524520874, "chaser_return": 19.75250244140625, "chaser_subspace_norm": 45.27481460571289, "collisions_per_episode": 18.274999618530273, "entropy": 0.11570126935839653, "explorer_grad_norm": 0.34453892707824707, "explorer_new_fields": 59.400001525878906, "explorer_partner_vision": 0.9137500524520874, "explorer_return": 24.787500381469727, "explorer_subspace_norm": 102.14765167236328, "final_distance": 1.780577540397644, "max_param_abs": 1.0218327045440674, "policy_loss": -0.0022482229445515145, "value_loss": 4.725666761398315}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:25:43.221561+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991543.2215755, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7497886683430666, "elapsed_seconds": 6299.239206035039, "env_steps_per_second": 10477.455743666336, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3711672932848929, "eta_seconds": 1336.2022558256144, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006575068350617268, "chaser_grad_norm": 0.9530724063515663, "chaser_new_fields": 51.150001525878906, "chaser_partner_vision": 0.906000018119812, "chaser_return": 20.177501678466797, "chaser_subspace_norm": 45.07564163208008, "collisions_per_episode": 18.725000381469727, "entropy": 0.18015236780047417, "explorer_grad_norm": 0.3635387420654297, "explorer_new_fields": 55.92499923706055, "explorer_partner_vision": 0.906000018119812, "explorer_return": 19.387500762939453, "explorer_subspace_norm": 90.44203186035156, "final_distance": 1.9510879516601562, "max_param_abs": 1.0339773893356323, "policy_loss": -0.003595833853818231, "value_loss": 5.0002933740615845}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:28:58.991813+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991738.991827, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8037059826194632, "elapsed_seconds": 6493.341537430068, "env_steps_per_second": 10472.266029442988, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31830105575637585, "eta_seconds": 1145.883800722953, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0009732602693413028, "chaser_grad_norm": 1.4392622858285904, "chaser_new_fields": 51.75, "chaser_partner_vision": 0.9160000681877136, "chaser_return": 17.1875, "chaser_subspace_norm": 46.5717658996582, "collisions_per_episode": 15.975000381469727, "entropy": 0.16272476315498352, "explorer_grad_norm": 0.5005936548113823, "explorer_new_fields": 61.650001525878906, "explorer_partner_vision": 0.9160000681877136, "explorer_return": 28.149999618530273, "explorer_subspace_norm": 92.76831817626953, "final_distance": 1.775112509727478, "max_param_abs": 1.0859367847442627, "policy_loss": -0.003525318900120311, "value_loss": 5.103221535682678}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:32:13.094145+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991933.0941591, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.8571440658050253, "elapsed_seconds": 6685.718636898091, "env_steps_per_second": 10470.078655968871, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26530629511500364, "eta_seconds": 955.102662414013, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0005849139146607563, "chaser_grad_norm": 1.2107825726270676, "chaser_new_fields": 48.57500076293945, "chaser_partner_vision": 0.9220000505447388, "chaser_return": 22.0, "chaser_subspace_norm": 46.91295623779297, "collisions_per_episode": 20.899999618530273, "entropy": 0.1707601547241211, "explorer_grad_norm": 0.33136148005723953, "explorer_new_fields": 57.25, "explorer_partner_vision": 0.9220000505447388, "explorer_return": 18.1875, "explorer_subspace_norm": 93.67933654785156, "final_distance": 1.8260315656661987, "max_param_abs": 1.0884759426116943, "policy_loss": -0.004376469029118368, "value_loss": 5.165245175361633}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:35:25.471243+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992125.4712548, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9099716799094717, "elapsed_seconds": 6875.898047674098, "env_steps_per_second": 10471.35945018198, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21221907554549685, "eta_seconds": 763.9886719637886, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0010855934286231328, "chaser_grad_norm": 0.2470189668238163, "chaser_new_fields": 49.72500228881836, "chaser_partner_vision": 0.877500057220459, "chaser_return": 21.100000381469727, "chaser_subspace_norm": 47.50455856323242, "collisions_per_episode": 19.899999618530273, "entropy": 0.16560807824134827, "explorer_grad_norm": 0.5236912816762924, "explorer_new_fields": 59.72500228881836, "explorer_partner_vision": 0.877500057220459, "explorer_return": 21.98750114440918, "explorer_subspace_norm": 101.11253356933594, "final_distance": 1.7961090803146362, "max_param_abs": 1.0859907865524292, "policy_loss": -0.00465683211479373, "value_loss": 5.161740064620972}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:38:35.650655+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992315.6506686, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9633579674055606, "elapsed_seconds": 7068.088682660018, "env_steps_per_second": 10469.591331181586, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1591911865463968, "eta_seconds": 573.0882715670285, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 1.3614476406242915e-05, "chaser_grad_norm": 0.8744113221764565, "chaser_new_fields": 50.47500228881836, "chaser_partner_vision": 0.92125004529953, "chaser_return": 23.9325008392334, "chaser_subspace_norm": 47.50455093383789, "collisions_per_episode": 22.42500114440918, "entropy": 0.15927444025874138, "explorer_grad_norm": 0.6268134489655495, "explorer_new_fields": 57.900001525878906, "explorer_partner_vision": 0.92125004529953, "explorer_return": 18.325000762939453, "explorer_subspace_norm": 99.32598114013672, "final_distance": 1.6602646112442017, "max_param_abs": 1.1491212844848633, "policy_loss": -0.003591384448576207, "value_loss": 5.220502257347107}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:41:47.841290+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992507.841303, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0168330304313957, "elapsed_seconds": 7260.598909553024, "env_steps_per_second": 10467.456052420708, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1061491068648103, "eta_seconds": 382.13678471331707, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0006092669887822755, "chaser_grad_norm": 0.35696154832839966, "chaser_new_fields": 45.775001525878906, "chaser_partner_vision": 0.8552500605583191, "chaser_return": 22.845001220703125, "chaser_subspace_norm": 46.37202072143555, "collisions_per_episode": 22.149999618530273, "entropy": 0.1844303160905838, "explorer_grad_norm": 0.2725492790341377, "explorer_new_fields": 54.525001525878906, "explorer_partner_vision": 0.8552500605583191, "explorer_return": 15.199999809265137, "explorer_subspace_norm": 109.85243225097656, "final_distance": 2.0060760974884033, "max_param_abs": 1.1527485847473145, "policy_loss": -0.0036055206032474807, "value_loss": 5.897579312324524}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:45:00.351517+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992700.3515308, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.068785811661413, "elapsed_seconds": 7447.628921981086, "env_steps_per_second": 10473.131894338772, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.05304579004260033, "eta_seconds": 190.96484415336118, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0004011438947220114, "chaser_grad_norm": 0.6709560602903366, "chaser_new_fields": 48.85000228881836, "chaser_partner_vision": 0.9065000414848328, "chaser_return": 18.16750144958496, "chaser_subspace_norm": 47.56572723388672, "collisions_per_episode": 17.42500114440918, "entropy": 0.15810470655560493, "explorer_grad_norm": 0.271642342209816, "explorer_new_fields": 61.45000076293945, "explorer_partner_vision": 0.9065000414848328, "explorer_return": 26.637500762939453, "explorer_subspace_norm": 130.63743591308594, "final_distance": 2.109950304031372, "max_param_abs": 1.1950421333312988, "policy_loss": -0.004216729789973783, "value_loss": 4.4083778858184814}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:48:07.381528+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992887.381539, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.106769434505, "elapsed_seconds": 7584.369964218, "env_steps_per_second": 10548.008651665048, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:50:24.122570+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993024.1225803, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.1067731915308587, "elapsed_seconds": 7584.383489511092, "env_steps_per_second": 10547.989841314971, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0022941239069333985, "chaser_grad_norm": 0.43133843690156937, "chaser_new_fields": 50.125, "chaser_partner_vision": 0.9205000400543213, "chaser_return": 20.852500915527344, "chaser_subspace_norm": 52.702186584472656, "collisions_per_episode": 19.67500114440918, "entropy": 0.15483620017766953, "explorer_grad_norm": 0.3490636646747589, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9205000400543213, "explorer_return": 23.0, "explorer_subspace_norm": 101.96681213378906, "final_distance": 1.8305482864379883, "max_param_abs": 1.1802449226379395, "policy_loss": -0.00493374926503698, "value_loss": 4.697651386260986}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:50:24.136096+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993024.136106, "update": 20000} diff --git a/social/seed_0008/attempt_01/status.json b/social/seed_0008/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..3c48befa518caa0e0c8c723ede4ff585cb08cb73 --- /dev/null +++ b/social/seed_0008/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.1067731915308587, + "elapsed_seconds": 7584.383489511092, + "env_steps_per_second": 10547.989841314971, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0022941239069333985, + "chaser_grad_norm": 0.43133843690156937, + "chaser_new_fields": 50.125, + "chaser_partner_vision": 0.9205000400543213, + "chaser_return": 20.852500915527344, + "chaser_subspace_norm": 52.702186584472656, + "collisions_per_episode": 19.67500114440918, + "entropy": 0.15483620017766953, + "explorer_grad_norm": 0.3490636646747589, + "explorer_new_fields": 60.525001525878906, + "explorer_partner_vision": 0.9205000400543213, + "explorer_return": 23.0, + "explorer_subspace_norm": 101.96681213378906, + "final_distance": 1.8305482864379883, + "max_param_abs": 1.1802449226379395, + "policy_loss": -0.00493374926503698, + "value_loss": 4.697651386260986 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:24.136096+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782993024.136106, + "update": 20000 +} diff --git a/social/seed_0009/attempt_01/checkpoints/latest.safetensors b/social/seed_0009/attempt_01/checkpoints/latest.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b8f6c095d062cf271d3974d40e61fb8c9ca2eb7b --- /dev/null +++ b/social/seed_0009/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:6790e80fe441a2995b3c9817361ed5add9b0fbc14ca34187b08685868aad6875 +size 11460376 diff --git a/social/seed_0009/attempt_01/config.json b/social/seed_0009/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..951f2904de26903146851e67f23051a8049b52ee --- /dev/null +++ b/social/seed_0009/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "mlp", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "checkpoint_every": 1000, + "checkpoint_every_seconds": 1800.0, + "clip_epsilon": 0.2, + "cost_per_hour": null, + "cuda_tf32": true, + "device": "cuda", + "entropy_coef": 0.01, + "env": { + "grid_size": 10, + "max_steps": 100, + "partner_visibility": "partial", + "task": "social", + "vision_radius": 3 + }, + "experiment_id": "mouse-run-run-2-mlp", + "finite_guard": true, + "fused_agent_rollout": false, + "gae_lambda": 0.95, + "gamma": 0.99, + "grad_clip": 1.0, + "hidden_size": 256, + "learning_rate": 0.0003, + "log_every": 500, + "metrics_path": null, + "ppo_epochs": 4, + "recurrent_l2_coef": 0.3, + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-mlp.social.seed_0009.attempt_01", + "seed": 9, + "status_every_seconds": 1800.0, + "status_path": null, + "subspace_metric_period": 10, + "tensorboard_dir": null, + "triton_env_step": true, + "updates": 20000, + "value_clip": 10.0, + "value_coef": 0.5 +} diff --git a/social/seed_0009/attempt_01/metrics.jsonl b/social/seed_0009/attempt_01/metrics.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..45fe9b2e06ef19d50eb53032c60f21caa22e85ab --- /dev/null +++ b/social/seed_0009/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0005420144574923648, "elapsed_seconds": 1.9512520469725132, "env_steps_per_second": 2049.9658187194445, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 10.839747135389803, "eta_seconds": 39023.08968740329, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00010244609255232427, "chaser_grad_norm": 0.18814831227064133, "chaser_new_fields": 34.47500228881836, "chaser_partner_vision": 0.3147500157356262, "chaser_return": -1.6175001859664917, "chaser_subspace_norm": 0.003209212562069297, "collisions_per_episode": 1.375, "entropy": 2.7724730372428894, "explorer_grad_norm": 0.11821835860610008, "explorer_new_fields": 35.47500228881836, "explorer_partner_vision": 0.3147500157356262, "explorer_return": 2.1500000953674316, "explorer_subspace_norm": 0.0030482904985547066, "final_distance": 4.796816349029541, "max_param_abs": 0.15428665280342102, "policy_loss": -0.000946801222512228, "value_loss": 3.0739704966545105}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:44:01.805504+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985441.8055215, "update": 1} +{"device": "cuda", "elapsed_hours": 0.05405996199694022, "elapsed_seconds": 194.6158631889848, "env_steps_per_second": 10276.654570844867, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.108338517880669, "eta_seconds": 7590.018664370407, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00018962832037150434, "chaser_grad_norm": 0.33221856132149696, "chaser_new_fields": 60.75, "chaser_partner_vision": 0.49250003695487976, "chaser_return": 8.672499656677246, "chaser_subspace_norm": 13.598347663879395, "collisions_per_episode": 6.125, "entropy": 0.6839510798454285, "explorer_grad_norm": 0.30588168650865555, "explorer_new_fields": 63.82500076293945, "explorer_partner_vision": 0.49250003695487976, "explorer_return": 40.76250076293945, "explorer_subspace_norm": 8.495759963989258, "final_distance": 4.033463954925537, "max_param_abs": 0.3071158528327942, "policy_loss": -0.0062216177014633445, "value_loss": 5.4333637952804565}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:47:14.470111+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985634.4701228, "update": 500} +{"device": "cuda", "elapsed_hours": 0.10634876348055615, "elapsed_seconds": 382.8555485300021, "env_steps_per_second": 10447.804701690366, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.020626506130567, "eta_seconds": 7274.25542207004, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0007498470010061098, "chaser_grad_norm": 0.48889319598674774, "chaser_new_fields": 53.150001525878906, "chaser_partner_vision": 0.6255000233650208, "chaser_return": 12.862500190734863, "chaser_subspace_norm": 26.601476669311523, "collisions_per_episode": 11.475000381469727, "entropy": 0.5308816879987717, "explorer_grad_norm": 0.23592163622379303, "explorer_new_fields": 62.22500228881836, "explorer_partner_vision": 0.6255000233650208, "explorer_return": 34.0, "explorer_subspace_norm": 9.286191940307617, "final_distance": 3.3776352405548096, "max_param_abs": 0.36255016922950745, "policy_loss": -0.006884164237888513, "value_loss": 5.897082567214966}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:50:22.709796+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782985822.7098086, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.16064761030830493, "elapsed_seconds": 578.3313971098978, "env_steps_per_second": 10374.67450320676, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.981320527135761, "eta_seconds": 7132.75389768874, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00035413269163075967, "chaser_grad_norm": 0.3701774403452873, "chaser_new_fields": 51.45000076293945, "chaser_partner_vision": 0.764750063419342, "chaser_return": 12.350000381469727, "chaser_subspace_norm": 33.56159973144531, "collisions_per_episode": 11.40000057220459, "entropy": 0.4545877128839493, "explorer_grad_norm": 0.24919547513127327, "explorer_new_fields": 63.275001525878906, "explorer_partner_vision": 0.764750063419342, "explorer_return": 35.53750228881836, "explorer_subspace_norm": 10.681936264038086, "final_distance": 2.8406848907470703, "max_param_abs": 0.42958471179008484, "policy_loss": -0.008022249411794746, "value_loss": 5.700997352600098}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:53:38.185645+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986018.1856575, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.21470751589387269, "elapsed_seconds": 772.9470572179416, "env_steps_per_second": 10349.997357897055, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9323676430448542, "eta_seconds": 6956.5235149614755, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002082567099597199, "chaser_grad_norm": 0.3255595304071903, "chaser_new_fields": 51.85000228881836, "chaser_partner_vision": 0.7460000514984131, "chaser_return": 11.315001487731934, "chaser_subspace_norm": 36.34425735473633, "collisions_per_episode": 10.350000381469727, "entropy": 0.3382515609264374, "explorer_grad_norm": 0.3072017878293991, "explorer_new_fields": 61.375, "explorer_partner_vision": 0.7460000514984131, "explorer_return": 33.70000076293945, "explorer_subspace_norm": 14.679262161254883, "final_distance": 2.176467180252075, "max_param_abs": 0.4797949194908142, "policy_loss": -0.005445840719623618, "value_loss": 5.547091245651245}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:56:52.801305+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986212.8013194, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.26634473813969334, "elapsed_seconds": 958.8410573028959, "env_steps_per_second": 10429.25719944533, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8644131669778532, "eta_seconds": 6711.8874011202715, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00043729556359872745, "chaser_grad_norm": 0.7258255034685135, "chaser_new_fields": 50.60000228881836, "chaser_partner_vision": 0.6720000505447388, "chaser_return": 13.417502403259277, "chaser_subspace_norm": 46.11403274536133, "collisions_per_episode": 12.475000381469727, "entropy": 0.2604081779718399, "explorer_grad_norm": 0.2999735698103905, "explorer_new_fields": 60.875, "explorer_partner_vision": 0.6720000505447388, "explorer_return": 31.73750114440918, "explorer_subspace_norm": 16.910747528076172, "final_distance": 2.4476864337921143, "max_param_abs": 0.4944987893104553, "policy_loss": -0.00428719109110709, "value_loss": 5.376678943634033}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T09:59:58.695304+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986398.695315, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.31969288640497123, "elapsed_seconds": 1150.8943910578964, "env_steps_per_second": 10426.673457822364, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8115930229615034, "eta_seconds": 6521.7348826614125, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00043213545042286583, "chaser_grad_norm": 0.30569563806056976, "chaser_new_fields": 46.025001525878906, "chaser_partner_vision": 0.8102500438690186, "chaser_return": 13.940001487731934, "chaser_subspace_norm": 51.89942169189453, "collisions_per_episode": 13.949999809265137, "entropy": 0.25289952009916306, "explorer_grad_norm": 0.2926001399755478, "explorer_new_fields": 61.650001525878906, "explorer_partner_vision": 0.8102500438690186, "explorer_return": 31.225000381469727, "explorer_subspace_norm": 18.942800521850586, "final_distance": 2.3614468574523926, "max_param_abs": 0.5005384087562561, "policy_loss": -0.007729922598223027, "value_loss": 5.507001638412476}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:03:10.748641+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986590.7486582, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.37218986226388046, "elapsed_seconds": 1339.8835041499697, "env_steps_per_second": 10448.669572122008, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7546093506725793, "eta_seconds": 6316.593662421285, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.00040487350064255523, "chaser_grad_norm": 1.086032211780548, "chaser_new_fields": 46.20000076293945, "chaser_partner_vision": 0.8997500538825989, "chaser_return": 20.072500228881836, "chaser_subspace_norm": 52.76332092285156, "collisions_per_episode": 19.625, "entropy": 0.21628035604953766, "explorer_grad_norm": 0.37785156816244125, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.8997500538825989, "explorer_return": 24.30000114440918, "explorer_subspace_norm": 24.48344612121582, "final_distance": 1.8529685735702515, "max_param_abs": 0.5352974534034729, "policy_loss": -0.004679075478573935, "value_loss": 5.83633816242218}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:06:19.737752+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986779.7377632, "update": 3500} +{"device": "cuda", "elapsed_hours": 0.42647396214054123, "elapsed_seconds": 1535.3062637059484, "env_steps_per_second": 10421.373492855379, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.705895848562165, "eta_seconds": 6141.2250548237935, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0005707659212284, "chaser_grad_norm": 0.7329047173261642, "chaser_new_fields": 50.025001525878906, "chaser_partner_vision": 0.921500027179718, "chaser_return": 19.122501373291016, "chaser_subspace_norm": 50.94835662841797, "collisions_per_episode": 18.075000762939453, "entropy": 0.19487255811691284, "explorer_grad_norm": 0.5545460432767868, "explorer_new_fields": 62.32500076293945, "explorer_partner_vision": 0.921500027179718, "explorer_return": 28.11250114440918, "explorer_subspace_norm": 25.17099952697754, "final_distance": 1.740604281425476, "max_param_abs": 0.5483516454696655, "policy_loss": -0.004662375320680923, "value_loss": 5.757978439331055}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:09:35.160512+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782986975.1605265, "update": 4000} +{"device": "cuda", "elapsed_hours": 0.4799525408772105, "elapsed_seconds": 1727.8291471579578, "env_steps_per_second": 10417.696697389052, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6531698630215028, "eta_seconds": 5951.41150687741, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0011784483821859482, "chaser_grad_norm": 0.3644195795059204, "chaser_new_fields": 46.0, "chaser_partner_vision": 0.8785000443458557, "chaser_return": 20.25749969482422, "chaser_subspace_norm": 48.10295867919922, "collisions_per_episode": 19.725000381469727, "entropy": 0.20208805426955223, "explorer_grad_norm": 0.29897354915738106, "explorer_new_fields": 59.45000076293945, "explorer_partner_vision": 0.8785000443458557, "explorer_return": 24.512500762939453, "explorer_subspace_norm": 27.591047286987305, "final_distance": 2.1820366382598877, "max_param_abs": 0.5756462216377258, "policy_loss": -0.004034561324871788, "value_loss": 6.039626598358154}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:12:47.683396+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987167.6834087, "update": 4500} +{"device": "cuda", "elapsed_hours": 0.5321639633872028, "elapsed_seconds": 1915.7902681939304, "env_steps_per_second": 10439.556110103098, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5964918901616088, "eta_seconds": 5747.370804581791, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0008088399920174183, "chaser_grad_norm": 0.20964409410953522, "chaser_new_fields": 48.0, "chaser_partner_vision": 0.9117500185966492, "chaser_return": 18.8075008392334, "chaser_subspace_norm": 46.64317321777344, "collisions_per_episode": 18.125, "entropy": 0.198086217045784, "explorer_grad_norm": 0.477555088698864, "explorer_new_fields": 60.92499923706055, "explorer_partner_vision": 0.9117500185966492, "explorer_return": 25.649999618530273, "explorer_subspace_norm": 30.35015869140625, "final_distance": 1.9205907583236694, "max_param_abs": 0.627726137638092, "policy_loss": -0.004872180681678273, "value_loss": 5.578304648399353}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:15:55.644516+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987355.64453, "update": 5000} +{"device": "cuda", "elapsed_hours": 0.5861748448766432, "elapsed_seconds": 2110.2294415559154, "env_steps_per_second": 10425.40662487343, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5453700455838772, "eta_seconds": 5563.332164101958, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005699657271556191, "chaser_grad_norm": 0.33466310799121857, "chaser_new_fields": 50.22500228881836, "chaser_partner_vision": 0.9030000567436218, "chaser_return": 17.0625, "chaser_subspace_norm": 43.615848541259766, "collisions_per_episode": 16.125, "entropy": 0.2043677493929863, "explorer_grad_norm": 0.3498583137989044, "explorer_new_fields": 64.95000457763672, "explorer_partner_vision": 0.9030000567436218, "explorer_return": 32.3125, "explorer_subspace_norm": 33.64988708496094, "final_distance": 1.8990494012832642, "max_param_abs": 0.6323936581611633, "policy_loss": -0.00510832616780732, "value_loss": 5.384105801582336}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:19:10.083692+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987550.083706, "update": 5500} +{"device": "cuda", "elapsed_hours": 0.6398453312744257, "elapsed_seconds": 2303.4431925879326, "env_steps_per_second": 10419.184669814173, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4929724396403266, "eta_seconds": 5374.700782705176, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0009387013047565951, "chaser_grad_norm": 0.4149181470274925, "chaser_new_fields": 49.625, "chaser_partner_vision": 0.9275000691413879, "chaser_return": 17.38250160217285, "chaser_subspace_norm": 44.8447151184082, "collisions_per_episode": 16.575000762939453, "entropy": 0.20752330124378204, "explorer_grad_norm": 0.2922320254147053, "explorer_new_fields": 63.42499923706055, "explorer_partner_vision": 0.9275000691413879, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 31.333335876464844, "final_distance": 1.7765525579452515, "max_param_abs": 0.6638135313987732, "policy_loss": -0.0058871528999873135, "value_loss": 5.410983085632324}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:22:23.297440+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987743.2974553, "update": 6000} +{"device": "cuda", "elapsed_hours": 0.6937267494752692, "elapsed_seconds": 2497.4162981109694, "env_steps_per_second": 10410.759319407918, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.440817095064021, "eta_seconds": 5186.9415422304755, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0014634896197511793, "chaser_grad_norm": 0.33461473137140274, "chaser_new_fields": 53.5, "chaser_partner_vision": 0.9257500171661377, "chaser_return": 16.747501373291016, "chaser_subspace_norm": 45.261112213134766, "collisions_per_episode": 15.324999809265137, "entropy": 0.2198631353676319, "explorer_grad_norm": 0.32219602912664413, "explorer_new_fields": 63.375, "explorer_partner_vision": 0.9257500171661377, "explorer_return": 31.73750114440918, "explorer_subspace_norm": 37.29536437988281, "final_distance": 1.9412078857421875, "max_param_abs": 0.6593057513237, "policy_loss": -0.0069273023847493675, "value_loss": 5.396772146224976}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:25:37.270544+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782987937.2705545, "update": 6500} +{"device": "cuda", "elapsed_hours": 0.7472298699044364, "elapsed_seconds": 2690.027531655971, "env_steps_per_second": 10408.815400771495, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3877126155368105, "eta_seconds": 4995.7654159325175, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.001261898741434564, "chaser_grad_norm": 0.3376742713153362, "chaser_new_fields": 52.400001525878906, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 16.255002975463867, "chaser_subspace_norm": 46.15476608276367, "collisions_per_episode": 15.050000190734863, "entropy": 0.19173558428883553, "explorer_grad_norm": 0.27360060065984726, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 26.512500762939453, "explorer_subspace_norm": 41.7067756652832, "final_distance": 1.8221057653427124, "max_param_abs": 0.6863431930541992, "policy_loss": -0.006027918285690603, "value_loss": 5.527910590171814}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:28:49.881779+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988129.881792, "update": 7000} +{"device": "cuda", "elapsed_hours": 0.7997698828741624, "elapsed_seconds": 2879.1715783469845, "env_steps_per_second": 10419.66384553708, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3329498047902704, "eta_seconds": 4798.6192972449735, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00016798326048927947, "chaser_grad_norm": 0.9336861670017242, "chaser_new_fields": 50.67499923706055, "chaser_partner_vision": 0.9230000376701355, "chaser_return": 21.010000228881836, "chaser_subspace_norm": 46.31756591796875, "collisions_per_episode": 19.75, "entropy": 0.1593332663178444, "explorer_grad_norm": 0.5640936493873596, "explorer_new_fields": 61.650001525878906, "explorer_partner_vision": 0.9230000376701355, "explorer_return": 26.337499618530273, "explorer_subspace_norm": 45.40749740600586, "final_distance": 1.7487242221832275, "max_param_abs": 0.7160663604736328, "policy_loss": -0.003416660249234482, "value_loss": 5.390875816345215}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:31:59.025825+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988319.0258367, "update": 7500} +{"device": "cuda", "elapsed_hours": 0.8529120313847024, "elapsed_seconds": 3070.4833129849285, "env_steps_per_second": 10421.812053064583, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2793680470770534, "eta_seconds": 4605.724969477393, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0008904790722192302, "chaser_grad_norm": 0.36171454191207886, "chaser_new_fields": 46.79999923706055, "chaser_partner_vision": 0.9362500309944153, "chaser_return": 18.575000762939453, "chaser_subspace_norm": 43.32550811767578, "collisions_per_episode": 18.25, "entropy": 0.21749116852879524, "explorer_grad_norm": 0.7838859111070633, "explorer_new_fields": 58.29999923706055, "explorer_partner_vision": 0.9362500309944153, "explorer_return": 21.5, "explorer_subspace_norm": 48.39027786254883, "final_distance": 1.6534041166305542, "max_param_abs": 0.7243714332580566, "policy_loss": -0.005512329521589976, "value_loss": 5.549964547157288}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:35:10.337562+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988510.3375754, "update": 8000} +{"device": "cuda", "elapsed_hours": 0.9055540315855372, "elapsed_seconds": 3259.994513707934, "env_steps_per_second": 10429.465404629847, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.225161336851021, "eta_seconds": 4410.580812663676, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00103654970721756, "chaser_grad_norm": 0.33552661910653114, "chaser_new_fields": 50.85000228881836, "chaser_partner_vision": 0.937000036239624, "chaser_return": 19.547500610351562, "chaser_subspace_norm": 43.39199447631836, "collisions_per_episode": 18.17500114440918, "entropy": 0.18336305022239685, "explorer_grad_norm": 0.24968143552541733, "explorer_new_fields": 58.875, "explorer_partner_vision": 0.937000036239624, "explorer_return": 23.1875, "explorer_subspace_norm": 54.92484664916992, "final_distance": 2.010357141494751, "max_param_abs": 0.7474641799926758, "policy_loss": -0.0041716138976681805, "value_loss": 5.536440014839172}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:38:19.848760+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988699.848772, "update": 8500} +{"device": "cuda", "elapsed_hours": 0.9581795166608127, "elapsed_seconds": 3449.446259978926, "env_steps_per_second": 10436.457705597055, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.1711082981409933, "eta_seconds": 4215.989873307576, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0009532780875902702, "chaser_grad_norm": 0.6602493114769459, "chaser_new_fields": 49.375, "chaser_partner_vision": 0.8962500691413879, "chaser_return": 20.755001068115234, "chaser_subspace_norm": 46.04595947265625, "collisions_per_episode": 19.700000762939453, "entropy": 0.2013452686369419, "explorer_grad_norm": 0.6484414339065552, "explorer_new_fields": 59.32500076293945, "explorer_partner_vision": 0.8962500691413879, "explorer_return": 23.8125, "explorer_subspace_norm": 58.37664794921875, "final_distance": 2.0456387996673584, "max_param_abs": 0.7762188911437988, "policy_loss": -0.005885201799213036, "value_loss": 6.205806851387024}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:41:29.300508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782988889.3005228, "update": 9000} +{"device": "cuda", "elapsed_hours": 1.0109849308035983, "elapsed_seconds": 3639.545750892954, "env_steps_per_second": 10440.863393646525, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.117404397203977, "eta_seconds": 4022.655829934318, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 1.028994244389736e-05, "chaser_grad_norm": 0.743528701364994, "chaser_new_fields": 39.67499923706055, "chaser_partner_vision": 0.9362500309944153, "chaser_return": 19.840002059936523, "chaser_subspace_norm": 60.863162994384766, "collisions_per_episode": 20.399999618530273, "entropy": 0.1486622914671898, "explorer_grad_norm": 0.3110369145870209, "explorer_new_fields": 46.650001525878906, "explorer_partner_vision": 0.9362500309944153, "explorer_return": 4.9375, "explorer_subspace_norm": 62.421165466308594, "final_distance": 1.387436866760254, "max_param_abs": 0.8006581664085388, "policy_loss": -0.0029016367157455036, "value_loss": 5.97783899307251}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:44:39.399997+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989079.4000094, "update": 9500} +{"device": "cuda", "elapsed_hours": 1.0641751730827511, "elapsed_seconds": 3831.030623097904, "env_steps_per_second": 10441.054623482652, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0641751730827511, "eta_seconds": 3831.030623097904, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -0.0005405637145701192, "chaser_grad_norm": 0.6586134396493435, "chaser_new_fields": 26.399999618530273, "chaser_partner_vision": 0.4767500162124634, "chaser_return": -0.964999794960022, "chaser_subspace_norm": 129.77442932128906, "collisions_per_episode": 3.6000001430511475, "entropy": 0.03203802462667227, "explorer_grad_norm": 1.0111998468637466, "explorer_new_fields": 70.75, "explorer_partner_vision": 0.4767500162124634, "explorer_return": 52.07500076293945, "explorer_subspace_norm": 108.2373046875, "final_distance": 5.926254749298096, "max_param_abs": 0.8170592188835144, "policy_loss": -0.0015848269728024889, "value_loss": 6.722829103469849}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:47:50.884870+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989270.8848848, "update": 10000} +{"device": "cuda", "elapsed_hours": 1.1158353211147025, "elapsed_seconds": 4017.007156012929, "env_steps_per_second": 10455.545227777737, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.00956529053235, "eta_seconds": 3634.43504591646, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0008793855390843985, "chaser_grad_norm": 0.6085812486708164, "chaser_new_fields": 47.92499923706055, "chaser_partner_vision": 0.9117500185966492, "chaser_return": 17.14000129699707, "chaser_subspace_norm": 58.89046859741211, "collisions_per_episode": 16.5, "entropy": 0.16924117133021355, "explorer_grad_norm": 0.6200921162962914, "explorer_new_fields": 60.625, "explorer_partner_vision": 0.9117500185966492, "explorer_return": 27.17500114440918, "explorer_subspace_norm": 79.057373046875, "final_distance": 1.7493332624435425, "max_param_abs": 0.8848351836204529, "policy_loss": -0.003300649800799249, "value_loss": 5.247099041938782}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:50:56.861403+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989456.8614135, "update": 10500} +{"device": "cuda", "elapsed_hours": 1.1681951553255527, "elapsed_seconds": 4205.50255917199, "env_steps_per_second": 10462.483230223748, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9557960361754522, "eta_seconds": 3440.865730231628, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.001397660440120907, "chaser_grad_norm": 0.3876924589276314, "chaser_new_fields": 47.0, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 20.23000144958496, "chaser_subspace_norm": 58.66804885864258, "collisions_per_episode": 19.649999618530273, "entropy": 0.15481824055314064, "explorer_grad_norm": 0.4604729861021042, "explorer_new_fields": 55.22500228881836, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 17.125, "explorer_subspace_norm": 66.17047119140625, "final_distance": 1.7231181859970093, "max_param_abs": 0.8849061131477356, "policy_loss": -0.005642088728031069, "value_loss": 5.0988664627075195}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:54:05.356807+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989645.3568208, "update": 11000} +{"device": "cuda", "elapsed_hours": 1.2214139331730418, "elapsed_seconds": 4397.090159422951, "env_steps_per_second": 10461.463907312007, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9027842114757266, "eta_seconds": 3250.0231613126157, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00030212402487533874, "chaser_grad_norm": 0.47057339549064636, "chaser_new_fields": 50.95000076293945, "chaser_partner_vision": 0.9550000429153442, "chaser_return": 20.052499771118164, "chaser_subspace_norm": 56.49590301513672, "collisions_per_episode": 18.825000762939453, "entropy": 0.13595974445343018, "explorer_grad_norm": 0.7706698030233383, "explorer_new_fields": 63.17499923706055, "explorer_partner_vision": 0.9550000429153442, "explorer_return": 30.0625, "explorer_subspace_norm": 77.67330169677734, "final_distance": 1.8816715478897095, "max_param_abs": 0.9543300867080688, "policy_loss": -0.003560056853572746, "value_loss": 4.980809330940247}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T10:57:16.944406+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782989836.9444187, "update": 11500} +{"device": "cuda", "elapsed_hours": 1.2757984051177562, "elapsed_seconds": 4592.874258423923, "env_steps_per_second": 10450.97194027505, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8505322700785042, "eta_seconds": 3061.916172282615, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00043596199381434886, "chaser_grad_norm": 0.3490357771515846, "chaser_new_fields": 47.375, "chaser_partner_vision": 0.9065000414848328, "chaser_return": 18.652502059936523, "chaser_subspace_norm": 54.14060592651367, "collisions_per_episode": 18.075000762939453, "entropy": 0.13214454799890518, "explorer_grad_norm": 0.5356516689062119, "explorer_new_fields": 58.25, "explorer_partner_vision": 0.9065000414848328, "explorer_return": 22.337499618530273, "explorer_subspace_norm": 74.10955810546875, "final_distance": 1.7980018854141235, "max_param_abs": 0.9477553963661194, "policy_loss": -0.00340679619554507, "value_loss": 4.1448341608047485}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:00:32.728505+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990032.728519, "update": 12000} +{"device": "cuda", "elapsed_hours": 1.3283241321508346, "elapsed_seconds": 4781.9668757430045, "env_steps_per_second": 10455.948629345365, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7969944792905006, "eta_seconds": 2869.1801254458023, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0010390090554590614, "chaser_grad_norm": 0.9886784106492996, "chaser_new_fields": 47.650001525878906, "chaser_partner_vision": 0.9422500729560852, "chaser_return": 20.642499923706055, "chaser_subspace_norm": 54.287567138671875, "collisions_per_episode": 19.975000381469727, "entropy": 0.14650164917111397, "explorer_grad_norm": 0.6492014154791832, "explorer_new_fields": 61.07500076293945, "explorer_partner_vision": 0.9422500729560852, "explorer_return": 24.537500381469727, "explorer_subspace_norm": 80.4127426147461, "final_distance": 1.726588487625122, "max_param_abs": 0.9545627236366272, "policy_loss": -0.004317821911070219, "value_loss": 4.642275810241699}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:03:41.821123+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990221.8211322, "update": 12500} +{"device": "cuda", "elapsed_hours": 1.3816188160905343, "elapsed_seconds": 4973.827737925923, "env_steps_per_second": 10454.72475926235, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.7439485932795185, "eta_seconds": 2678.2149358062666, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -7.32375198233548e-05, "chaser_grad_norm": 0.6503089293837547, "chaser_new_fields": 49.775001525878906, "chaser_partner_vision": 0.9100000262260437, "chaser_return": 20.89750099182129, "chaser_subspace_norm": 60.41501235961914, "collisions_per_episode": 19.625, "entropy": 0.12950463593006134, "explorer_grad_norm": 0.4018133729696274, "explorer_new_fields": 58.32500076293945, "explorer_partner_vision": 0.9100000262260437, "explorer_return": 21.825000762939453, "explorer_subspace_norm": 83.03009796142578, "final_distance": 1.6191853284835815, "max_param_abs": 0.9776861071586609, "policy_loss": -0.003792193079831918, "value_loss": 5.183174133300781}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:06:53.681988+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990413.6820035, "update": 13000} +{"device": "cuda", "elapsed_hours": 1.4349016682547193, "elapsed_seconds": 5165.64600571699, "env_steps_per_second": 10453.677998886573, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6908785810115315, "eta_seconds": 2487.1628916415134, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -2.795188086324174e-06, "chaser_grad_norm": 0.4167756214737892, "chaser_new_fields": 47.45000076293945, "chaser_partner_vision": 0.9142500162124634, "chaser_return": 27.61749839782715, "chaser_subspace_norm": 52.125526428222656, "collisions_per_episode": 26.375, "entropy": 0.1492275893688202, "explorer_grad_norm": 0.26703913882374763, "explorer_new_fields": 57.07500076293945, "explorer_partner_vision": 0.9142500162124634, "explorer_return": 16.200000762939453, "explorer_subspace_norm": 86.04097747802734, "final_distance": 1.5821764469146729, "max_param_abs": 0.990484893321991, "policy_loss": -0.004047125945797703, "value_loss": 5.7425795793533325}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:10:05.500252+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990605.5002642, "update": 13500} +{"device": "cuda", "elapsed_hours": 1.4874028961016383, "elapsed_seconds": 5354.650425965898, "env_steps_per_second": 10458.199050388701, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6374583840435593, "eta_seconds": 2294.8501825568133, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0010306610292394325, "chaser_grad_norm": 0.27848588302731514, "chaser_new_fields": 51.20000076293945, "chaser_partner_vision": 0.9232500195503235, "chaser_return": 20.00750160217285, "chaser_subspace_norm": 54.040462493896484, "collisions_per_episode": 18.575000762939453, "entropy": 0.1652975007891655, "explorer_grad_norm": 0.30660809576511383, "explorer_new_fields": 59.875, "explorer_partner_vision": 0.9232500195503235, "explorer_return": 24.712499618530273, "explorer_subspace_norm": 87.00945281982422, "final_distance": 1.6969026327133179, "max_param_abs": 0.9625483751296997, "policy_loss": -0.005140351034700785, "value_loss": 4.892420768737793}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:13:14.504674+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990794.5046878, "update": 14000} +{"device": "cuda", "elapsed_hours": 1.5398490670438818, "elapsed_seconds": 5543.456641357974, "env_steps_per_second": 10462.785902803022, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5840806806028517, "eta_seconds": 2102.690450170266, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0004954629318363324, "chaser_grad_norm": 0.7749978750944138, "chaser_new_fields": 48.79999923706055, "chaser_partner_vision": 0.9110000729560852, "chaser_return": 22.18000030517578, "chaser_subspace_norm": 55.45209884643555, "collisions_per_episode": 21.05000114440918, "entropy": 0.12484971806406975, "explorer_grad_norm": 0.4433329105377197, "explorer_new_fields": 63.10000228881836, "explorer_partner_vision": 0.9110000729560852, "explorer_return": 27.412500381469727, "explorer_subspace_norm": 117.27658081054688, "final_distance": 2.0648865699768066, "max_param_abs": 1.0853676795959473, "policy_loss": -0.003968824634327461, "value_loss": 4.935833692550659}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:16:23.310888+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782990983.3108995, "update": 14500} +{"device": "cuda", "elapsed_hours": 1.5931688034058122, "elapsed_seconds": 5735.407692260924, "env_steps_per_second": 10461.331298376756, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5310562678019374, "eta_seconds": 1911.8025640869746, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0008720776557380461, "chaser_grad_norm": 0.3638540208339691, "chaser_new_fields": 47.525001525878906, "chaser_partner_vision": 0.9412500262260437, "chaser_return": 15.532501220703125, "chaser_subspace_norm": 49.01909637451172, "collisions_per_episode": 15.125, "entropy": 0.17501505464315414, "explorer_grad_norm": 0.2871592566370964, "explorer_new_fields": 59.47500228881836, "explorer_partner_vision": 0.9412500262260437, "explorer_return": 25.875, "explorer_subspace_norm": 98.15377044677734, "final_distance": 1.7637592554092407, "max_param_abs": 1.07170832157135, "policy_loss": -0.00473757475288572, "value_loss": 5.138084769248962}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:19:35.261942+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991175.2619572, "update": 15000} +{"device": "cuda", "elapsed_hours": 1.6465781443999732, "elapsed_seconds": 5927.681319839903, "env_steps_per_second": 10459.401687551333, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.47803881611612126, "eta_seconds": 1720.9397380180365, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0007403472015248269, "chaser_grad_norm": 0.35511621460318565, "chaser_new_fields": 47.025001525878906, "chaser_partner_vision": 0.9315000176429749, "chaser_return": 21.892501831054688, "chaser_subspace_norm": 51.96568298339844, "collisions_per_episode": 20.975000381469727, "entropy": 0.14311188086867332, "explorer_grad_norm": 0.5615191459655762, "explorer_new_fields": 61.32500076293945, "explorer_partner_vision": 0.9315000176429749, "explorer_return": 22.80000114440918, "explorer_subspace_norm": 108.31959533691406, "final_distance": 1.709702491760254, "max_param_abs": 1.0703400373458862, "policy_loss": -0.003508009449575411, "value_loss": 4.876586318016052}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:22:47.535567+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991367.5355775, "update": 15500} +{"device": "cuda", "elapsed_hours": 1.6992875940772096, "elapsed_seconds": 6117.435338677955, "env_steps_per_second": 10461.900528045648, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4248218985193024, "eta_seconds": 1529.3588346694887, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0012445409460970502, "chaser_grad_norm": 0.6611562892794609, "chaser_new_fields": 49.82500076293945, "chaser_partner_vision": 0.9320000410079956, "chaser_return": 18.877500534057617, "chaser_subspace_norm": 51.89919662475586, "collisions_per_episode": 17.67500114440918, "entropy": 0.18040956184267998, "explorer_grad_norm": 0.26360610127449036, "explorer_new_fields": 55.54999923706055, "explorer_partner_vision": 0.9320000410079956, "explorer_return": 18.337499618530273, "explorer_subspace_norm": 83.97859191894531, "final_distance": 1.6843055486679077, "max_param_abs": 1.0703400373458862, "policy_loss": -0.004156645252372471, "value_loss": 4.762564659118652}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:25:57.289587+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991557.289603, "update": 16000} +{"device": "cuda", "elapsed_hours": 1.7533453339241614, "elapsed_seconds": 6312.043202126981, "env_steps_per_second": 10456.20219737404, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.3719217374990645, "eta_seconds": 1338.9182549966322, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0009054262759633502, "chaser_grad_norm": 0.612411618232727, "chaser_new_fields": 51.72500228881836, "chaser_partner_vision": 0.9350000619888306, "chaser_return": 22.715002059936523, "chaser_subspace_norm": 59.58201599121094, "collisions_per_episode": 21.100000381469727, "entropy": 0.11782051436603069, "explorer_grad_norm": 0.38480208069086075, "explorer_new_fields": 58.70000076293945, "explorer_partner_vision": 0.9350000619888306, "explorer_return": 19.212499618530273, "explorer_subspace_norm": 119.71931457519531, "final_distance": 1.958505630493164, "max_param_abs": 1.0703400373458862, "policy_loss": -0.0039037133995449835, "value_loss": 4.645286798477173}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:29:11.897449+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991751.8974595, "update": 16500} +{"device": "cuda", "elapsed_hours": 1.8056362762974782, "elapsed_seconds": 6500.290594670922, "env_steps_per_second": 10461.070779781425, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.31864169581720203, "eta_seconds": 1147.1101049419274, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0005389512174422961, "chaser_grad_norm": 0.5940982028841972, "chaser_new_fields": 50.875, "chaser_partner_vision": 0.9297500252723694, "chaser_return": 20.127500534057617, "chaser_subspace_norm": 51.4810676574707, "collisions_per_episode": 18.725000381469727, "entropy": 0.13607750833034515, "explorer_grad_norm": 0.24371851980686188, "explorer_new_fields": 59.79999923706055, "explorer_partner_vision": 0.9297500252723694, "explorer_return": 22.950000762939453, "explorer_subspace_norm": 122.69559478759766, "final_distance": 1.7938969135284424, "max_param_abs": 1.099482774734497, "policy_loss": -0.003080795256884983, "value_loss": 4.76505172252655}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:32:20.144842+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782991940.144856, "update": 17000} +{"device": "cuda", "elapsed_hours": 1.857619064014709, "elapsed_seconds": 6687.428630452952, "env_steps_per_second": 10467.401428590463, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2653741520021013, "eta_seconds": 955.3469472075647, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0014228231096699018, "chaser_grad_norm": 0.212687436491251, "chaser_new_fields": 52.57500076293945, "chaser_partner_vision": 0.8880000710487366, "chaser_return": 20.42500114440918, "chaser_subspace_norm": 55.26328659057617, "collisions_per_episode": 18.80000114440918, "entropy": 0.13055263087153435, "explorer_grad_norm": 0.537092000246048, "explorer_new_fields": 61.17499923706055, "explorer_partner_vision": 0.8880000710487366, "explorer_return": 26.212499618530273, "explorer_subspace_norm": 125.14385223388672, "final_distance": 1.848523736000061, "max_param_abs": 1.0998934507369995, "policy_loss": -0.0030579660539515885, "value_loss": 4.952346563339233}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:35:27.282877+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992127.2828875, "update": 17500} +{"device": "cuda", "elapsed_hours": 1.9119024077088882, "elapsed_seconds": 6882.848667751998, "env_steps_per_second": 10460.784985341814, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.21243360085654311, "eta_seconds": 764.7609630835552, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0014346631403139298, "chaser_grad_norm": 0.5796013176441193, "chaser_new_fields": 48.025001525878906, "chaser_partner_vision": 0.924250066280365, "chaser_return": 19.097501754760742, "chaser_subspace_norm": 53.05559539794922, "collisions_per_episode": 18.325000762939453, "entropy": 0.17466740682721138, "explorer_grad_norm": 0.2960214838385582, "explorer_new_fields": 58.72500228881836, "explorer_partner_vision": 0.924250066280365, "explorer_return": 23.0, "explorer_subspace_norm": 142.1593475341797, "final_distance": 1.9546196460723877, "max_param_abs": 1.1035313606262207, "policy_loss": -0.004966565399431744, "value_loss": 5.409593224525452}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:38:42.702916+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992322.70293, "update": 18000} +{"device": "cuda", "elapsed_hours": 1.9640888912158294, "elapsed_seconds": 7070.720008376986, "env_steps_per_second": 10465.695136043998, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1592504506391213, "eta_seconds": 573.3016223008367, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0008777035390263865, "chaser_grad_norm": 0.2407803237438202, "chaser_new_fields": 49.275001525878906, "chaser_partner_vision": 0.9430000185966492, "chaser_return": 20.352500915527344, "chaser_subspace_norm": 55.509979248046875, "collisions_per_episode": 19.325000762939453, "entropy": 0.12164635770022869, "explorer_grad_norm": 0.5441446900367737, "explorer_new_fields": 55.775001525878906, "explorer_partner_vision": 0.9430000185966492, "explorer_return": 16.837499618530273, "explorer_subspace_norm": 111.63951873779297, "final_distance": 1.6061385869979858, "max_param_abs": 1.1012747287750244, "policy_loss": -0.003085129293614486, "value_loss": 4.171186208724976}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:41:50.574255+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992510.5742664, "update": 18500} +{"device": "cuda", "elapsed_hours": 2.0172136881672, "elapsed_seconds": 7261.96927740192, "env_steps_per_second": 10465.480794100818, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.10616914148248421, "eta_seconds": 382.2089093369432, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.000712290955394046, "chaser_grad_norm": 0.29661698639392853, "chaser_new_fields": 49.775001525878906, "chaser_partner_vision": 0.9395000338554382, "chaser_return": 18.942502975463867, "chaser_subspace_norm": 50.39433670043945, "collisions_per_episode": 18.024999618530273, "entropy": 0.20037353783845901, "explorer_grad_norm": 0.2258487045764923, "explorer_new_fields": 58.79999923706055, "explorer_partner_vision": 0.9395000338554382, "explorer_return": 22.700000762939453, "explorer_subspace_norm": 110.66793060302734, "final_distance": 1.7465057373046875, "max_param_abs": 1.140512466430664, "policy_loss": -0.004796268612146548, "value_loss": 5.0214524269104}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:45:01.823525+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992701.823539, "update": 19000} +{"device": "cuda", "elapsed_hours": 2.0703854436566407, "elapsed_seconds": 7453.387597163906, "env_steps_per_second": 10465.040088573931, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.053086806247606165, "eta_seconds": 191.1125024913822, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00014525146205501827, "chaser_grad_norm": 0.3561606705188751, "chaser_new_fields": 50.17499923706055, "chaser_partner_vision": 0.937000036239624, "chaser_return": 19.137502670288086, "chaser_subspace_norm": 53.63298797607422, "collisions_per_episode": 18.225000381469727, "entropy": 0.1071413941681385, "explorer_grad_norm": 0.727757565677166, "explorer_new_fields": 57.85000228881836, "explorer_partner_vision": 0.937000036239624, "explorer_return": 21.962499618530273, "explorer_subspace_norm": 120.7088623046875, "final_distance": 1.5730633735656738, "max_param_abs": 1.2036426067352295, "policy_loss": -0.001656085483031866, "value_loss": 4.085147738456726}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:48:13.241843+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782992893.2418542, "update": 19500} +{"device": "cuda", "elapsed_hours": 2.1071196418872247, "elapsed_seconds": 7585.630710794008, "env_steps_per_second": 10546.255552114293, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T11:50:25.484957+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993025.484966, "update": 20000} +{"device": "cuda", "elapsed_hours": 2.107122335001089, "elapsed_seconds": 7585.64040600392, "env_steps_per_second": 10546.242072941028, "environment_steps": 80000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.0, "eta_seconds": 0.0, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": -0.0002899035942244277, "chaser_grad_norm": 0.2534610331058502, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 22.76500129699707, "chaser_subspace_norm": 57.69636535644531, "collisions_per_episode": 21.25, "entropy": 0.1398237645626068, "explorer_grad_norm": 0.3390047997236252, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 21.6875, "explorer_subspace_norm": 105.33492279052734, "final_distance": 1.5442389249801636, "max_param_abs": 1.1976157426834106, "policy_loss": -0.002912267684005876, "value_loss": 4.631349444389343}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T11:50:25.494653+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1782993025.494663, "update": 20000} diff --git a/social/seed_0009/attempt_01/status.json b/social/seed_0009/attempt_01/status.json new file mode 100644 index 0000000000000000000000000000000000000000..a6b75909b8a78f69395e407a6741110470513cf6 --- /dev/null +++ b/social/seed_0009/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 2.107122335001089, + "elapsed_seconds": 7585.64040600392, + "env_steps_per_second": 10546.242072941028, + "environment_steps": 80000000, + "error": null, + "estimated_cost_usd": null, + "eta_hours": 0.0, + "eta_seconds": 0.0, + "healthy": true, + "latest_checkpoint": "runs/mouse-run-run-2-mlp/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": -0.0002899035942244277, + "chaser_grad_norm": 0.2534610331058502, + "chaser_new_fields": 51.625, + "chaser_partner_vision": 0.9552500247955322, + "chaser_return": 22.76500129699707, + "chaser_subspace_norm": 57.69636535644531, + "collisions_per_episode": 21.25, + "entropy": 0.1398237645626068, + "explorer_grad_norm": 0.3390047997236252, + "explorer_new_fields": 58.42499923706055, + "explorer_partner_vision": 0.9552500247955322, + "explorer_return": 21.6875, + "explorer_subspace_norm": 105.33492279052734, + "final_distance": 1.5442389249801636, + "max_param_abs": 1.1976157426834106, + "policy_loss": -0.002912267684005876, + "value_loss": 4.631349444389343 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T11:50:25.494653+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1782993025.494663, + "update": 20000 +} diff --git a/tables.log b/tables.log new file mode 100644 index 0000000000000000000000000000000000000000..e3e90f1807c65594f9f7d0d7d7d768972203fbe6 --- /dev/null +++ b/tables.log @@ -0,0 +1,67 @@ +{ + "experiment_root": "runs/mouse-run-run-2-mlp", + "generated_at": "2026-07-02T11:51:00.748915+00:00", + "output_hashes": { + "evaluations": "265ed5550a98149da5384ac6cbcd551db5b121176f7046a4ca58d89531a77cc7", + "exclusions": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "rollouts": "140147549967098f1a78dadab8e5888eb41a37389c094f79b623d22d7f0bd74a", + "runs": "bfab762e5dd3673a8e08fa06635ebebd07bc02783f016fc26e56fee41756e2d8" + }, + "outputs": { + "evaluations": "runs/tables/mouse-run-run-2-mlp/evaluations.jsonl", + "exclusions": "runs/tables/mouse-run-run-2-mlp/exclusions.jsonl", + "rollouts": "runs/tables/mouse-run-run-2-mlp/rollouts.jsonl", + "runs": "runs/tables/mouse-run-run-2-mlp/runs.jsonl" + }, + "panel_path": "runs/tables/mouse-run-run-2-mlp/panels.json", + "provenance": { + "command": "scripts/transform/build_tables.py runs/mouse-run-run-2-mlp", + "created_at": "2026-07-02T11:51:00.749193+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "row_counts": { + "evaluations": 40, + "exclusions": 0, + "rollouts": 20, + "runs": 20 + }, + "schema_version": 1, + "transform_script": "/root/mouse-run-run-perf/scripts/transform/build_tables.py" +} diff --git a/triton_equivalence.json b/triton_equivalence.json new file mode 100644 index 0000000000000000000000000000000000000000..9e7bc875fb8f765c418bb32a655b4eea3c5f0afb --- /dev/null +++ b/triton_equivalence.json @@ -0,0 +1,69 @@ +{ + "batch_size": 40, + "cases": [ + { + "partner_visibility": "partial", + "task": "social" + }, + { + "partner_visibility": "none", + "task": "non_social" + }, + { + "partner_visibility": "full", + "task": "social" + } + ], + "command": "scripts/verify_triton_env.py --output runs/mouse-run-run-2-mlp/triton_equivalence.json", + "completed_at": "2026-07-02T07:36:14.368125+00:00", + "device": "cuda", + "grid_size": 10, + "provenance": { + "command": "scripts/verify_triton_env.py --output runs/mouse-run-run-2-mlp/triton_equivalence.json", + "created_at": "2026-07-02T07:36:14.368153+00:00", + "dependencies": { + "huggingface_hub": "1.21.0", + "numpy": "2.5.0", + "safetensors": "0.8.0", + "tensorboard": "2.21.0", + "torch": "2.12.1", + "triton": "3.7.1" + }, + "executable": "/root/mouse-run-run-perf/.venv/bin/python3", + "git": { + "branch": "main", + "diff_sha256": "e3b0c44298fc1c149afbf4c8996fb92427ae41e4649b934ca495991b7852b855", + "dirty": false, + "sha": "6282c650d281b31d646d537dd3f57c968806a423", + "short_sha": "6282c65", + "status_short": "" + }, + "platform": "Linux-6.8.0-117-generic-x86_64-with-glibc2.39", + "python": "3.14.5 (main, Jun 2 2026, 22:25:17) [Clang 22.1.3 ]", + "schema_version": 1, + "torch": { + "cuda_available": true, + "cuda_device_count": 1, + "cuda_devices": [ + { + "index": 0, + "major": 8, + "minor": 9, + "multi_processor_count": 66, + "name": "NVIDIA GeForce RTX 4070 Ti SUPER", + "total_memory_bytes": 16717578240 + } + ], + "cuda_version": "13.0", + "cudnn_version": 92000, + "mps_available": false, + "version": "2.12.1+cu130" + } + }, + "schema_version": 1, + "seeds": 20, + "started_at": "2026-07-02T07:36:06.084150+00:00", + "status": "ok", + "steps": 100, + "vision_radius": 3 +}