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..717f7537e78e5104bb791ef50d1379ddb0fa40ff --- /dev/null +++ b/HF_UPLOAD_MANIFEST.json @@ -0,0 +1,674 @@ +{ + "artifact_path": "runs/mouse-run-run-2-transformer", + "created_at": "2026-07-03T19:19:48.555969+00:00", + "excluded_from_file_manifest": [ + "runs/mouse-run-run-2-transformer/HF_UPLOAD_MANIFEST.json" + ], + "file_count": 125, + "files": [ + { + "path": "README.md", + "sha256": "02b919c0be5fbbfd2975aaa0accfdeb9350ecc300ef282256972ed5b2b569d0b", + "size_bytes": 2904 + }, + { + "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": "646d5089aeff3e68f08e4df142aae32fe3eaae03e4d6c7c5b1accc3b1751d913", + "size_bytes": 2782 + }, + { + "path": "derived/reports/figures/evaluation_comparison.png", + "sha256": "738e6a7547a9273692572f73c129ed2efb88ce00f18cfe1cd45f901a1235df4f", + "size_bytes": 111943 + }, + { + "path": "derived/reports/figures/training_curves.png", + "sha256": "9b4efbf32536bd83ad3bf2bfac46159b0cf15cc6d83acfd03820807246092357", + "size_bytes": 343071 + }, + { + "path": "derived/reports/report.md", + "sha256": "07ea6e1146a6d4391c08ed0d3763306a39917b879cc982c34946a4f2d9750cbd", + "size_bytes": 3017 + }, + { + "path": "derived/reports/summary.json", + "sha256": "7c1b08c438b29c8e14d9d9720f4cd9aa3e080f2aaa96b541b62351a9301a09d6", + "size_bytes": 14897 + }, + { + "path": "derived/tables/MANIFEST.json", + "sha256": "da05a5245daba06f81c15a1d2825867050fbb101cd80d8f0a879f95157bcf3a2", + "size_bytes": 2384 + }, + { + "path": "derived/tables/evaluations.jsonl", + "sha256": "8de65701cca4a6d5b3b6ec45d68a8bccf77698027577760bc3ac20c4f70ad50f", + "size_bytes": 139093 + }, + { + "path": "derived/tables/exclusions.jsonl", + "sha256": "31ff7680a00ea1dfa23d4aaa86ec7118d8706edf2005365152a3d680f8a314e1", + "size_bytes": 3663 + }, + { + "path": "derived/tables/panels.json", + "sha256": "cb73e84a077cbcd2d639c1c7f6947f3a418fa7482b62d1b789d5ad9608d30562", + "size_bytes": 760 + }, + { + "path": "derived/tables/rollouts.jsonl", + "sha256": "c521dba8fba745966a2515f4198f58eec0e2d6e265ccc164f823d5f04b5cdc3b", + "size_bytes": 24314 + }, + { + "path": "derived/tables/runs.jsonl", + "sha256": "bc8443607b1eacf8afb9ec05470609ad4c5c39d1f50d4ce78ff520465f3426d6", + "size_bytes": 50500 + }, + { + "path": "launch_gate.json", + "sha256": "0311afcdda4a0d9924e172fc0383e09f1a761124372d9e13999d37a60c2fe521", + "size_bytes": 4802 + }, + { + "path": "manifest.json", + "sha256": "2ac15af023b115bba05d12a98696623261af96c3cadd1936598c15d14869eeca", + "size_bytes": 15186 + }, + { + "path": "non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "sha256": "bebcbee7c71f93593b985b9403dddac3f4b9d301088feea80adec365e5c839d3", + "size_bytes": 29760368 + }, + { + "path": "non_social/seed_0000/attempt_01/config.json", + "sha256": "3886076e3107b0a4f5a93e3ad03817d35e2c0ce09d49a016d49bec2aa3877385", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0000/attempt_01/metrics.jsonl", + "sha256": "75912192efe4be72c477cc7de037d5e9df4a6e7a89cb9c75e9a4f965b1e51868", + "size_bytes": 54458 + }, + { + "path": "non_social/seed_0000/attempt_01/status.json", + "sha256": "8c66b9b3f5e2cb8af6d728d5486ca9b09a6c998bffd9cd2250871f31f134ec07", + "size_bytes": 1393 + }, + { + "path": "non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "sha256": "97e55e51803fcb1f12344ef9426ee05617a44dd6420294c15ba2d71524feac2c", + "size_bytes": 29760368 + }, + { + "path": "non_social/seed_0001/attempt_01/config.json", + "sha256": "d024907cb6c1ada7871a6aa3200527ec41df6b356ab615ef9be8e466e6c9f540", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0001/attempt_01/metrics.jsonl", + "sha256": "c77de502df9a63405ebf7483cf79c6a3b19f5956ae76bdb76b4e8ad731c7d08f", + "size_bytes": 54435 + }, + { + "path": "non_social/seed_0001/attempt_01/status.json", + "sha256": "db65d653e314dbb5f0c78a0d353dcda94e56c97f4fc3d12e1f56494f8c073028", + "size_bytes": 1389 + }, + { + "path": "non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "sha256": "6b05e3f4ca6b8f514ca68a393b1bdc046876c2fdb9517fb51b9942361d3a57ac", + "size_bytes": 29760368 + }, + { + "path": "non_social/seed_0002/attempt_01/config.json", + "sha256": "86dadf6a81b16adcd9d8c5b49178175a60848362f59fb2f4248aaaf8f0be73b8", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0002/attempt_01/metrics.jsonl", + "sha256": "abd38be484fbcf3287052a7e04ba148b97fbe5441c7760b7c2bc4ce10084fe3f", + "size_bytes": 54545 + }, + { + "path": "non_social/seed_0002/attempt_01/status.json", + "sha256": "3da5035da36a2c3aea567515cd5893a766b08e1fabe80389b8ff0bccccde1fbc", + "size_bytes": 1397 + }, + { + "path": "non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "sha256": "a06bb3351d710015bd4527f14150baa0ed7be3ea4e957deb0bac2ad812840a89", + "size_bytes": 29760352 + }, + { + "path": "non_social/seed_0003/attempt_01/config.json", + "sha256": "58774ca21b63655b8b786e8d3aacc1d76f3fa7aef96888600c429bbe1d558c05", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0003/attempt_01/metrics.jsonl", + "sha256": "9efff90342274c9f9134bb9142940fef6fdf04431c413a5fe6ff3f0d6a3879aa", + "size_bytes": 54335 + }, + { + "path": "non_social/seed_0003/attempt_01/status.json", + "sha256": "ffd17a46f134f1144e851364b021ec18bf47a2178d033e289c563a817fb0644a", + "size_bytes": 1376 + }, + { + "path": "non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "sha256": "7fc0051ad27e22e8e395faf9f846864d859748404592f14a524f3596f1f93559", + "size_bytes": 29760360 + }, + { + "path": "non_social/seed_0004/attempt_01/config.json", + "sha256": "2c87c1b048a5640816c623fa4dd03a45a5e69a1644a987d8d8c9b42df12c2153", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0004/attempt_01/metrics.jsonl", + "sha256": "dd2ff87c4c056c9b7babc70781047e108f6cbafeff2ad5b06ad6c8894c7eb881", + "size_bytes": 54469 + }, + { + "path": "non_social/seed_0004/attempt_01/status.json", + "sha256": "7a1c2a675f244d0d37c3379d7e7907ac4998fba181070e84dae6d2b2041b112c", + "size_bytes": 1382 + }, + { + "path": "non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "sha256": "ac8c8ce4aab7af2d19fc15778486415942ca2cc3a405b65f8c5b23c7cfb69376", + "size_bytes": 29760360 + }, + { + "path": "non_social/seed_0005/attempt_01/config.json", + "sha256": "26fcacf461990235094e53773c91fcc8cf1f2495b4b8ced68e9088c7d91a1952", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0005/attempt_01/metrics.jsonl", + "sha256": "05f4bcab7222f1ad1aaa622e855d60c274bc2876a19a96c4bbec3fc65c00fcf8", + "size_bytes": 54666 + }, + { + "path": "non_social/seed_0005/attempt_01/status.json", + "sha256": "57db486373f69de367239a75f4d2327dd0ac14b2c6582982a675c6646e785dc1", + "size_bytes": 1384 + }, + { + "path": "non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "sha256": "c3eb7fb2c91b2c029361c2a8a5a023ab29aba5b6cae2aca1cd33fac54dfe1d6a", + "size_bytes": 29760368 + }, + { + "path": "non_social/seed_0006/attempt_01/config.json", + "sha256": "614faee0db30f57d9d2aee9fec5a10ba51bd15ad50b6ea500633615e240d412c", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0006/attempt_01/metrics.jsonl", + "sha256": "c376069bfc0b0097ee1a1dd3d193cfe9f5e78024861cf4fd4302903878ab1501", + "size_bytes": 54548 + }, + { + "path": "non_social/seed_0006/attempt_01/status.json", + "sha256": "135a01d6bb28e9a97e7737b39b2d80e6e8936bc6481951076ab918b4b384f3ec", + "size_bytes": 1394 + }, + { + "path": "non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "sha256": "021e19e4992a272e4a09a90a4bb4bbedf2703d0d2a2122af4b17a128246da307", + "size_bytes": 29760368 + }, + { + "path": "non_social/seed_0007/attempt_01/config.json", + "sha256": "9bb3289524f274d1bf4efe1bbd94504490615579360be843e0307a9b6c07fbd4", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0007/attempt_01/metrics.jsonl", + "sha256": "5c14924f66958cffe3728c0c7ccd721dbbe68f842897f3bc501831bf655cdd45", + "size_bytes": 54589 + }, + { + "path": "non_social/seed_0007/attempt_01/status.json", + "sha256": "089bf762f103acfc6608f77b3d1bc982dcb4fc284a162e82ae0b26b1461d6529", + "size_bytes": 1396 + }, + { + "path": "non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "sha256": "8bd159d34635da80bf0bdbc948924eb30df6571f77a8dc53b6d138adad560333", + "size_bytes": 29760336 + }, + { + "path": "non_social/seed_0008/attempt_01/config.json", + "sha256": "d09109c22646e93c0c4dde6c32d7ecc6ec9155f8df6f422b3b77b828676f0a70", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0008/attempt_01/metrics.jsonl", + "sha256": "b1525ff219fe287b430e9b4c105641edd78090a90db29a24e538ee2c9122801e", + "size_bytes": 54465 + }, + { + "path": "non_social/seed_0008/attempt_01/status.json", + "sha256": "7c5f288999d30389b6cae79e66880601588e00105372d09fa125801512b6c023", + "size_bytes": 1361 + }, + { + "path": "non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "sha256": "6c3ce75ac26f6605968ebb1613d72e3f3e59ac898be34d14b34a59812bd25a49", + "size_bytes": 29760344 + }, + { + "path": "non_social/seed_0009/attempt_01/config.json", + "sha256": "b24d20d38c13df5a094b09cbdc16c8b4cba523d784c1baceff42c8210869e1dc", + "size_bytes": 1225 + }, + { + "path": "non_social/seed_0009/attempt_01/metrics.jsonl", + "sha256": "9fc047903def1bc2c246da61dc8a10ede51440f6654610a56edf41505e73532a", + "size_bytes": 54511 + }, + { + "path": "non_social/seed_0009/attempt_01/status.json", + "sha256": "fb931a5143914c38d6ab32d4a2a7ebedc1e565345fde56d044a02bc70b12c4d0", + "size_bytes": 1369 + }, + { + "path": "paper_random_eval.jsonl", + "sha256": "d78f6495e5f264d00f8b8bf084c877871b52a88d58d55efdb57de7219a6b5827", + "size_bytes": 135342 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors", + "sha256": "79523bf93d01ed70b7e868b54bd70459c210b81fefc7e06836c12becd787418b", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors", + "sha256": "4a525f188e61798e466bfab07962c00ba3a32419728ab58c132854d5c15021a6", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors", + "sha256": "1a3ffcae843364d6a41194d3fa074e3e38567f536a4c773bfce6d18933683747", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors", + "sha256": "fdff410f6f4486b5465b7d7e9b0eea876e81ec84939a13d65c18966e634c34f7", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors", + "sha256": "8a30011c7be774107fd2f2ec56a1411497995644a45e40ae9042e69fe31b8a7d", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors", + "sha256": "331de7d69aa252a1e6c25f85652611cfe8c65e76dc2d65502eea41dba4370686", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors", + "sha256": "49399764286403a10871d4a229bf4c66cc34b98d57dbf844493500003cff4e74", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors", + "sha256": "bdd9bfaf3009fd86f8b51becb610eac87ef0e29815d79f8a036fc7c78907441a", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors", + "sha256": "c46462eb90ada39824c0444c9e0be4903798ac5be414d9cf2d6edf098253bb45", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors", + "sha256": "2f95b021dd1d9ea44cf0be2a67663e3e14a847b79713c15072c335dd1ee53f51", + "size_bytes": 46506715 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors", + "sha256": "7b692d07d996c19666bcec9d3cbd1ca49895bbdaf265f987065c8c0a07a5bf67", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors", + "sha256": "5c982d42a21462cc56fbabd3d2e3cd0d0367cf46aad3e7751355630c4c16a448", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors", + "sha256": "fbd829a179d3930a2e2023754d754bfcd5da946d6a364fee2c4c15764e81cf5a", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors", + "sha256": "7850ba0b8693746b2fc891d045f690281838a55ff69f0ca4e7b0792ffb600089", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors", + "sha256": "4d30fd2cf0d2050b4e36b40486d11328ffb2f6223eb365ed5d24adebf7a042d6", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors", + "sha256": "61cab96377113bf461cc9b98496f068e46843299576259856285e4c5173396e4", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors", + "sha256": "fd6b3b52c19b03cbe1bc347b56180787b658f72ee1d28efeddae9a403e23ca1d", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors", + "sha256": "7ab97b49dbf08bc614e87382b28a0987ed96f4d5d3033b8b522164c2433de0fe", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors", + "sha256": "efa669fe4caed034302ce55a56e212b2cc51232ab9140ebeaf18711ad69dabb0", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors", + "sha256": "148864f819405af7554f128fa3574e5d75c0bb6afc957d717e24cb5752490e52", + "size_bytes": 46506699 + }, + { + "path": "paper_rollouts/paper_rollout_records.jsonl", + "sha256": "4d176c4e2f7a0290740ed4ca9122e1949522b03b9ef01cbf931c1f0b3c51186e", + "size_bytes": 21020 + }, + { + "path": "raw_records.jsonl", + "sha256": "445504330bfebe2c30711d6ef170368f4dae3cb3e677aa1e5d3babea4d64b75c", + "size_bytes": 47050 + }, + { + "path": "run_status.json", + "sha256": "230f229245a5e2d6608d281605d77ca32e9c0dcacb1e1da86fd1ebd862c0434f", + "size_bytes": 39468 + }, + { + "path": "social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "sha256": "d43d74b10a625e6385b2bb95ea1788823a3302c478fc62063d73fa7275a2e64b", + "size_bytes": 29760376 + }, + { + "path": "social/seed_0000/attempt_01/config.json", + "sha256": "ac913cc7582a6ea11146af201067c0dd0ca6039bd301a218690bb9a758fb8964", + "size_bytes": 1212 + }, + { + "path": "social/seed_0000/attempt_01/metrics.jsonl", + "sha256": "94cf22269f9a2350ef0bed9daeb3318e24ce6e6145a6717f5d66f2eaea04df1a", + "size_bytes": 55395 + }, + { + "path": "social/seed_0000/attempt_01/status.json", + "sha256": "33b0214be2971b51f3db1c5e4beb4cdc23473c62b1a462d2a45e1dfeedb19f1e", + "size_bytes": 1405 + }, + { + "path": "social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "sha256": "f65e21859c2db411274fb48eff4489b878589dc01d994df540c2fd45a29d7070", + "size_bytes": 29760384 + }, + { + "path": "social/seed_0001/attempt_01/config.json", + "sha256": "cdfbf77007da5b1e1056324dbec5d24a5c12fd430751c774c86e1481d1d46591", + "size_bytes": 1212 + }, + { + "path": "social/seed_0001/attempt_01/metrics.jsonl", + "sha256": "58bcc923be8e7341a9f6f333c05f80e2084e6e53a61db9aa0a2fe9bd663abe23", + "size_bytes": 55504 + }, + { + "path": "social/seed_0001/attempt_01/status.json", + "sha256": "cf5cfaba4653eeafec5b0d11eafda608a4ce5f6470e8afe3b701b2ae822a5261", + "size_bytes": 1412 + }, + { + "path": "social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "sha256": "c1f4144d32c140593489ce5522c50515a469ce6c2ad5aae7e6125078757e8ada", + "size_bytes": 29760368 + }, + { + "path": "social/seed_0002/attempt_01/config.json", + "sha256": "338e7d109047f61bec737052cf8b8c3debd4dad66d084e7954c6acb2b408eda3", + "size_bytes": 1212 + }, + { + "path": "social/seed_0002/attempt_01/metrics.jsonl", + "sha256": "e440afee038286e728ce19d5af9dccc9245dd0c9bc5c2a2a9e0b6a9ec3a71d89", + "size_bytes": 55488 + }, + { + "path": "social/seed_0002/attempt_01/status.json", + "sha256": "c0acbd148b7eef4c26f411b298bb22e8b18bafe0d0676eb9c96ab82a4ce3a5ee", + "size_bytes": 1402 + }, + { + "path": "social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "sha256": "42e883b23fb099df0bbca6fff98d93505336beb9d24fcca6844a6a86809443c2", + "size_bytes": 29760368 + }, + { + "path": "social/seed_0003/attempt_01/config.json", + "sha256": "1cd7b200d980d1f437677a0f8991dda2cd44ed80de15471f229978ff4b525cfd", + "size_bytes": 1212 + }, + { + "path": "social/seed_0003/attempt_01/metrics.jsonl", + "sha256": "18f56873b94f1b2fc72534fa3a30bf0a609123a225d66aa537eb3f363768bb02", + "size_bytes": 55429 + }, + { + "path": "social/seed_0003/attempt_01/status.json", + "sha256": "6c5d14007b34fe6cae16e2fea4787421ab9f57e03b294edb1b677126f4792958", + "size_bytes": 1402 + }, + { + "path": "social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "sha256": "b8b54f8a46f42267d8050df70d059bed5fffe88df8f02d91f48faba250143759", + "size_bytes": 29760376 + }, + { + "path": "social/seed_0004/attempt_01/config.json", + "sha256": "17a097f5a6d0b4e9367511937201418d7678682f862ef8b14402fee30a5ff36b", + "size_bytes": 1212 + }, + { + "path": "social/seed_0004/attempt_01/metrics.jsonl", + "sha256": "01f78b3d1e1c5ee35bd544ec1e52ee4ab68821f35a137d5fca80ce8d1d32556c", + "size_bytes": 55539 + }, + { + "path": "social/seed_0004/attempt_01/status.json", + "sha256": "1d405e52a854336c00db8ea295214df34ede4d7c05485c83ea11e7dd647cb2d3", + "size_bytes": 1412 + }, + { + "path": "social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "sha256": "78eaccb338b45f6a03f515f2cee3766c9898f3ac6218f1e0839ec5175bd738c0", + "size_bytes": 29760344 + }, + { + "path": "social/seed_0005/attempt_01/config.json", + "sha256": "d3e7cd3414bdfd674fc783ef961336fb3b354f83777630b5a45dac92ed308778", + "size_bytes": 1212 + }, + { + "path": "social/seed_0005/attempt_01/metrics.jsonl", + "sha256": "5986539aec56bfe2a4e152631cfe863bc2f3cc08d6a24480d0447fc1507b4022", + "size_bytes": 55334 + }, + { + "path": "social/seed_0005/attempt_01/status.json", + "sha256": "cb0e5e507cde74f70ea86364c004728496426b3af162ba3e6f80be9797e4c398", + "size_bytes": 1383 + }, + { + "path": "social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "sha256": "23e8a61977d52ecaa6a6bca53b80271a5b31024659be373ca295584a0cf5e3fa", + "size_bytes": 29760368 + }, + { + "path": "social/seed_0006/attempt_01/config.json", + "sha256": "ab1cd032b5fd5a4944b480e17fef6949b97e25f979aed5999c1f7098b0aa1dc5", + "size_bytes": 1212 + }, + { + "path": "social/seed_0006/attempt_01/metrics.jsonl", + "sha256": "a0ebcd73df03367aba224d567299249a175f00bec4632eb4da47c56107252638", + "size_bytes": 55460 + }, + { + "path": "social/seed_0006/attempt_01/status.json", + "sha256": "5bdb94a91f3f1e423934673f24dd47f6f9ff53fc76ad77055bf46717b673ad7c", + "size_bytes": 1403 + }, + { + "path": "social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "sha256": "31976e8959949de1246f21e9902b4882a9b47763a935458165086ad52ff66214", + "size_bytes": 29760376 + }, + { + "path": "social/seed_0007/attempt_01/config.json", + "sha256": "5514fc1f1cf4be651fd36a5027306de97959e429f343f3f0b1784918b6f6c040", + "size_bytes": 1212 + }, + { + "path": "social/seed_0007/attempt_01/metrics.jsonl", + "sha256": "160836bcdb07381ae668df54079a14604f1bbbb1b501068ebb12430c907aed44", + "size_bytes": 55456 + }, + { + "path": "social/seed_0007/attempt_01/status.json", + "sha256": "dafa1372eb70529ea83a90bcbe746a48bd2a400771c62dc1415d71f1f006a3c8", + "size_bytes": 1411 + }, + { + "path": "social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "sha256": "ff3f0e0f69d3eb842d30e1463bfb424b5f279fa52d10afb3cf854b7b57747b13", + "size_bytes": 29760384 + }, + { + "path": "social/seed_0008/attempt_01/config.json", + "sha256": "922725c0461eec84ba313fce07fc98c90ec2b09e5b4a4ced872b3560fe5fd828", + "size_bytes": 1212 + }, + { + "path": "social/seed_0008/attempt_01/metrics.jsonl", + "sha256": "82a71aa50ab724d0aa3f6427a14627ed360e81074e179d3a78b208a11c7b28f5", + "size_bytes": 55464 + }, + { + "path": "social/seed_0008/attempt_01/status.json", + "sha256": "c6d62b1b85f2e595a71c6e408c79ee37baf0533bc8d48000cdc91ee74c23a3e8", + "size_bytes": 1417 + }, + { + "path": "social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "sha256": "d50b2b6c17101b2b486d954abdd7be83fe3678f1e69fd0fffeec0607f1fc7685", + "size_bytes": 29760368 + }, + { + "path": "social/seed_0009/attempt_01/config.json", + "sha256": "352697e7d6177a28a3e8281abc1ede3a27c3cdaae9dea03ef57906d7dc8ca40e", + "size_bytes": 1212 + }, + { + "path": "social/seed_0009/attempt_01/metrics.jsonl", + "sha256": "14993be488100bc62c64edad70f025737e9490b7761ec961bc7b99ad071e4f10", + "size_bytes": 55398 + }, + { + "path": "social/seed_0009/attempt_01/status.json", + "sha256": "19c1aacf7ae3caeba46581065da684a98425ee41d0e30b5632908a34938a1da0", + "size_bytes": 1401 + }, + { + "path": "triton_equivalence.json", + "sha256": "6c51db69737bc2cc7696492f10ad3c661f0f0add505c8b7153e9a609403d3ed2", + "size_bytes": 1951 + } + ], + "path_in_repo": "", + "private": false, + "provenance": { + "command": "/Users/jacoblincool/Documents/GitHub/mouse-run-run/.venv/bin/upload-marl-artifacts runs/mouse-run-run-2-transformer --repo-id JacobLinCool/mouse-run-run-2-transformer --repo-type model --commit-message 'Upload mouse-run-run-2-transformer: trained agents, rollouts, evaluation, cross-architecture study'", + "created_at": "2026-07-03T19:19:49.136304+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-transformer", + "repo_type": "model", + "revision": null, + "schema_version": 1 +} diff --git a/README.md b/README.md new file mode 100644 index 0000000000000000000000000000000000000000..e7ffb62a353a697c236761fa988dc8cd295fa43d --- /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 — TRANSFORMER agents (mouse-run-run-2-transformer) + +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:** 2-layer causal transformer (no evolving state). 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: **41.31** + (non-social control: 3.39) +- chaser partner-in-vision: 0.96 + +Neural (social agents): +- collision decoding (balanced acc): **0.85** +- partner escape / approach decoding: 0.76 / 0.76 +- PLSC shared-dimension top correlation: 0.81 + +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-transformer", + "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..f3a778c53492bae921110d39e2ecf30fec3e5350 --- /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": "2026-07-03T01:24:40.542114+00:00", + "input_hashes": { + "evaluations.jsonl": "8de65701cca4a6d5b3b6ec45d68a8bccf77698027577760bc3ac20c4f70ad50f", + "exclusions.jsonl": "31ff7680a00ea1dfa23d4aaa86ec7118d8706edf2005365152a3d680f8a314e1", + "rollouts.jsonl": "c521dba8fba745966a2515f4198f58eec0e2d6e265ccc164f823d5f04b5cdc3b", + "runs.jsonl": "bc8443607b1eacf8afb9ec05470609ad4c5c39d1f50d4ce78ff520465f3426d6" + }, + "input_tables": { + "evaluations.jsonl": "runs/tables/mouse-run-run-2-transformer/evaluations.jsonl", + "exclusions.jsonl": "runs/tables/mouse-run-run-2-transformer/exclusions.jsonl", + "rollouts.jsonl": "runs/tables/mouse-run-run-2-transformer/rollouts.jsonl", + "runs.jsonl": "runs/tables/mouse-run-run-2-transformer/runs.jsonl" + }, + "output_hashes": { + "evaluation_comparison": "738e6a7547a9273692572f73c129ed2efb88ce00f18cfe1cd45f901a1235df4f", + "training_curves": "9b4efbf32536bd83ad3bf2bfac46159b0cf15cc6d83acfd03820807246092357" + }, + "outputs": { + "figures": [ + "runs/reports/mouse-run-run-2-transformer/figures/training_curves.png", + "runs/reports/mouse-run-run-2-transformer/figures/evaluation_comparison.png" + ], + "report": "runs/reports/mouse-run-run-2-transformer/report.md", + "summary": "runs/reports/mouse-run-run-2-transformer/summary.json" + }, + "provenance": { + "command": "scripts/analysis/summarize_experiment.py runs/tables/mouse-run-run-2-transformer", + "created_at": "2026-07-03T01:24:40.542642+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..5656166d1b23069c1071f09d72e08d3457d61bb5 --- /dev/null +++ b/derived/reports/figures/evaluation_comparison.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:738e6a7547a9273692572f73c129ed2efb88ce00f18cfe1cd45f901a1235df4f +size 111943 diff --git a/derived/reports/figures/training_curves.png b/derived/reports/figures/training_curves.png new file mode 100644 index 0000000000000000000000000000000000000000..a8813395cea3f7d810733c950cd4933f3c2bed99 --- /dev/null +++ b/derived/reports/figures/training_curves.png @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:9b4efbf32536bd83ad3bf2bfac46159b0cf15cc6d83acfd03820807246092357 +size 343071 diff --git a/derived/reports/report.md b/derived/reports/report.md new file mode 100644 index 0000000000000000000000000000000000000000..a7aec76a9a615c147cb607ab479395aeaa57da93 --- /dev/null +++ b/derived/reports/report.md @@ -0,0 +1,88 @@ +# MARL Experiment Summary + +Generated: `2026-07-03T01:24:40.539994+00:00` +Tables: `runs/tables/mouse-run-run-2-transformer` + +## 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 | 0.748 ± 0.267 | 10.745 ± 2.13 | +| chaser return | 8.787 ± 0.193 | 12.937 ± 2.166 | +| explorer return | 94.041 ± 1.533 | 41.021 ± 5.261 | +| chaser partner vision | 0 ± 0 | 0.97 ± 0.01 | +| explorer partner vision | 0 ± 0 | 0.97 ± 0.01 | +| chaser new fields | 94.165 ± 0.998 | 58.885 ± 2.309 | +| explorer new fields | 96.328 ± 1.137 | 66.545 ± 2.836 | +| final distance | 3.248 ± 0.764 | 1.673 ± 0.117 | +| policy loss | -0.000632 ± 0.000994 | -0.005 ± 0.001 | +| value loss | 0.08 ± 0.047 | 2.879 ± 0.469 | +| entropy | 0.002 ± 0.000794 | 0.154 ± 0.027 | +| approx kl | 0.000827 ± 0.001 | 0.003 ± 0.001 | +| chaser grad norm | 0.058 ± 0.06 | 0.704 ± 0.362 | +| explorer grad norm | 0.311 ± 0.15 | 0.957 ± 0.34 | +| max param abs | 2.849 ± 0.52 | 2.752 ± 0.159 | + +## Random-Opponent Evaluation + +### Opponent: `random_chaser` + +| Metric | `non_social` (n=10) | `social` (n=10) | +| --- | ---: | ---: | +| collisions per episode | 2.797 ± 0.255 | 0.277 ± 0.052 | +| chaser return | -3.25 ± 0.046 | -2.903 ± 0.061 | +| explorer return | 90.611 ± 1.929 | 6.096 ± 7.816 | +| chaser partner vision | 0 ± 0 | 0.163 ± 0.021 | +| explorer partner vision | 0 ± 0 | 0.163 ± 0.021 | +| chaser new fields | 34.021 ± 0.216 | 34.03 ± 0.058 | +| explorer new fields | 94.393 ± 1.267 | 37.601 ± 5.236 | +| final distance | 4.085 ± 0.228 | 6.707 ± 0.182 | +| average distance | 5.08 ± 0.029 | 6.528 ± 0.209 | +| degenerate fraction | 0.879 ± 0.018 | 0.96 ± 0.02 | + +### Opponent: `random_explorer` + +| Metric | `non_social` (n=10) | `social` (n=10) | +| --- | ---: | ---: | +| collisions per episode | 3.395 ± 0.471 | 41.308 ± 3.288 | +| chaser return | 8.349 ± 0.203 | 40.052 ± 3.481 | +| explorer return | -1.149 ± 0.219 | -43.318 ± 3.18 | +| chaser partner vision | 0 ± 0 | 0.961 ± 0.007 | +| explorer partner vision | 0 ± 0 | 0.961 ± 0.007 | +| chaser new fields | 92.85 ± 0.907 | 26.842 ± 1.072 | +| explorer new fields | 33.097 ± 0.088 | 23.326 ± 1.153 | +| final distance | 4.388 ± 0.329 | 1.403 ± 0.083 | +| average distance | 5.111 ± 0.029 | 1.571 ± 0.052 | +| degenerate fraction | 0.853 ± 0.014 | 0.982 ± 0.015 | + +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 | +| ---: | ---: | ---: | +| 59 | 300 | 12 | + +## Failed Attempts +- count: 0 diff --git a/derived/reports/summary.json b/derived/reports/summary.json new file mode 100644 index 0000000000000000000000000000000000000000..9186d53b8c4cccc7e4d80b22659e0bd50eccd675 --- /dev/null +++ b/derived/reports/summary.json @@ -0,0 +1,544 @@ +{ + "evaluation_metrics": { + "failed_count": 0, + "non_social": { + "random_chaser": { + "average_distance": { + "max": 5.129418849945068, + "mean": 5.0796972751617435, + "min": 5.043961048126221, + "n": 10, + "std": 0.02873995140186124 + }, + "chaser_new_fields": { + "max": 34.36000061035156, + "mean": 34.020999908447266, + "min": 33.72999954223633, + "n": 10, + "std": 0.21644684003085007 + }, + "chaser_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "chaser_return": { + "max": -3.1859986782073975, + "mean": -3.2501988649368285, + "min": -3.3139989376068115, + "n": 10, + "std": 0.04577726762110156 + }, + "collisions_per_episode": { + "max": 3.059999942779541, + "mean": 2.796999955177307, + "min": 2.319999933242798, + "n": 10, + "std": 0.2553840545416018 + }, + "degenerate_fraction": { + "max": 0.91, + "mean": 0.8789999999999999, + "min": 0.86, + "n": 10, + "std": 0.01813835714721707 + }, + "explorer_new_fields": { + "max": 96.38999938964844, + "mean": 94.39299774169922, + "min": 92.62999725341797, + "n": 10, + "std": 1.267440183318429 + }, + "explorer_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "explorer_return": { + "max": 93.50499725341797, + "mean": 90.61149673461914, + "min": 87.93999481201172, + "n": 10, + "std": 1.9294984069099073 + }, + "final_distance": { + "max": 4.635434627532959, + "mean": 4.085444545745849, + "min": 3.837200880050659, + "n": 10, + "std": 0.22845185221708492 + }, + "n": 10 + }, + "random_explorer": { + "average_distance": { + "max": 5.154397487640381, + "mean": 5.11141185760498, + "min": 5.0569047927856445, + "n": 10, + "std": 0.028723257963978274 + }, + "chaser_new_fields": { + "max": 93.91999816894531, + "mean": 92.8499969482422, + "min": 90.43000030517578, + "n": 10, + "std": 0.9074680661154696 + }, + "chaser_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "chaser_return": { + "max": 8.579997062683105, + "mean": 8.349396276474, + "min": 7.805994510650635, + "n": 10, + "std": 0.20263148293859806 + }, + "collisions_per_episode": { + "max": 4.210000038146973, + "mean": 3.394999933242798, + "min": 2.6599998474121094, + "n": 10, + "std": 0.4708556415010507 + }, + "degenerate_fraction": { + "max": 0.87, + "mean": 0.853, + "min": 0.82, + "n": 10, + "std": 0.014177446878757837 + }, + "explorer_new_fields": { + "max": 33.21999740600586, + "mean": 33.09699897766113, + "min": 32.939998626708984, + "n": 10, + "std": 0.08809638710169938 + }, + "explorer_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "explorer_return": { + "max": -0.8899999856948853, + "mean": -1.149499958753586, + "min": -1.59499990940094, + "n": 10, + "std": 0.21923101569002812 + }, + "final_distance": { + "max": 5.0355939865112305, + "mean": 4.388345527648926, + "min": 4.020629405975342, + "n": 10, + "std": 0.3294317958141699 + }, + "n": 10 + } + }, + "social": { + "random_chaser": { + "average_distance": { + "max": 6.819681167602539, + "mean": 6.528078365325928, + "min": 6.161418914794922, + "n": 10, + "std": 0.20855749231374748 + }, + "chaser_new_fields": { + "max": 34.11000061035156, + "mean": 34.029999542236325, + "min": 33.939998626708984, + "n": 10, + "std": 0.057620230332741544 + }, + "chaser_partner_vision": { + "max": 0.20059996843338013, + "mean": 0.163489992916584, + "min": 0.13729999959468842, + "n": 10, + "std": 0.021343070859131442 + }, + "chaser_return": { + "max": -2.780999183654785, + "mean": -2.9034990072250366, + "min": -2.978998899459839, + "n": 10, + "std": 0.06094620981282601 + }, + "collisions_per_episode": { + "max": 0.3700000047683716, + "mean": 0.27699999809265136, + "min": 0.20999999344348907, + "n": 10, + "std": 0.05158488367146182 + }, + "degenerate_fraction": { + "max": 1.0, + "mean": 0.96, + "min": 0.93, + "n": 10, + "std": 0.020493901531919198 + }, + "explorer_new_fields": { + "max": 46.5, + "mean": 37.600999450683595, + "min": 30.35999870300293, + "n": 10, + "std": 5.235516137790574 + }, + "explorer_partner_vision": { + "max": 0.20059996843338013, + "mean": 0.163489992916584, + "min": 0.13729999959468842, + "n": 10, + "std": 0.021343070859131442 + }, + "explorer_return": { + "max": 19.404998779296875, + "mean": 6.096499764919281, + "min": -4.670000076293945, + "n": 10, + "std": 7.816309965617632 + }, + "final_distance": { + "max": 6.965867042541504, + "mean": 6.706823682785034, + "min": 6.317955017089844, + "n": 10, + "std": 0.18157252453346814 + }, + "n": 10 + }, + "random_explorer": { + "average_distance": { + "max": 1.667794942855835, + "mean": 1.5712934255599975, + "min": 1.4947947263717651, + "n": 10, + "std": 0.05184039584079168 + }, + "chaser_new_fields": { + "max": 28.6299991607666, + "mean": 26.841999435424803, + "min": 25.19999885559082, + "n": 10, + "std": 1.0724253811717708 + }, + "chaser_partner_vision": { + "max": 0.9711999893188477, + "mean": 0.9609299838542938, + "min": 0.9440000057220459, + "n": 10, + "std": 0.006767867998884705 + }, + "chaser_return": { + "max": 44.46000289916992, + "mean": 40.052001571655275, + "min": 34.46500015258789, + "n": 10, + "std": 3.4813239901279833 + }, + "collisions_per_episode": { + "max": 45.36000061035156, + "mean": 41.30799903869629, + "min": 36.189998626708984, + "n": 10, + "std": 3.2882451909664763 + }, + "degenerate_fraction": { + "max": 1.0, + "mean": 0.982, + "min": 0.96, + "n": 10, + "std": 0.01536229149573723 + }, + "explorer_new_fields": { + "max": 25.18000030517578, + "mean": 23.325999641418456, + "min": 21.5, + "n": 10, + "std": 1.1533272603517895 + }, + "explorer_partner_vision": { + "max": 0.9711999893188477, + "mean": 0.9609299838542938, + "min": 0.9440000057220459, + "n": 10, + "std": 0.006767867998884705 + }, + "explorer_return": { + "max": -38.09000015258789, + "mean": -43.3179988861084, + "min": -46.93000030517578, + "n": 10, + "std": 3.1799977794472576 + }, + "final_distance": { + "max": 1.5478826761245728, + "mean": 1.403444504737854, + "min": 1.2944831848144531, + "n": 10, + "std": 0.08294436722867861 + }, + "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-transformer/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-transformer/figures/evaluation_comparison.png" + } + ], + "generated_at": "2026-07-03T01:24:40.539994+00:00", + "rollout_exclusions": { + "episodes_with_exclusion_records": 300, + "excluded_degenerate_episodes": 59, + "records": 12 + }, + "schema_version": 2, + "tables_root": "runs/tables/mouse-run-run-2-transformer", + "training_metrics": { + "non_social": { + "approx_kl": { + "max": 0.0031268233263772593, + "mean": 0.0008274657820166141, + "min": 3.0426282704221208e-05, + "n": 10, + "std": 0.0011119185808850796 + }, + "chaser_grad_norm": { + "max": 0.2260504849255085, + "mean": 0.05761161218397319, + "min": 0.02223971695639193, + "n": 10, + "std": 0.06012663616499861 + }, + "chaser_new_fields": { + "max": 95.80000305175781, + "mean": 94.1650016784668, + "min": 91.82500457763672, + "n": 10, + "std": 0.9980102893568585 + }, + "chaser_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "chaser_return": { + "max": 9.135000228881836, + "mean": 8.787000560760498, + "min": 8.360000610351562, + "n": 10, + "std": 0.19291447540213363 + }, + "collisions_per_episode": { + "max": 1.149999976158142, + "mean": 0.7475000083446502, + "min": 0.3499999940395355, + "n": 10, + "std": 0.2667981449728105 + }, + "entropy": { + "max": 0.003422852954827249, + "mean": 0.0021465113706653936, + "min": 0.0009539475431665778, + "n": 10, + "std": 0.0007943372788127037 + }, + "explorer_grad_norm": { + "max": 0.6812433376908302, + "mean": 0.31067706923931837, + "min": 0.1489477176219225, + "n": 10, + "std": 0.15010061774843952 + }, + "explorer_new_fields": { + "max": 97.95000457763672, + "mean": 96.32750244140625, + "min": 94.82500457763672, + "n": 10, + "std": 1.1369062620540298 + }, + "explorer_partner_vision": { + "max": 0.0, + "mean": 0.0, + "min": 0.0, + "n": 10, + "std": 0.0 + }, + "explorer_return": { + "max": 96.0625, + "mean": 94.04125213623047, + "min": 91.82500457763672, + "n": 10, + "std": 1.5333274673425037 + }, + "final_distance": { + "max": 4.808191776275635, + "mean": 3.247923755645752, + "min": 2.2879505157470703, + "n": 10, + "std": 0.7642974377711885 + }, + "max_param_abs": { + "max": 3.969390392303467, + "mean": 2.849128818511963, + "min": 2.185072422027588, + "n": 10, + "std": 0.5202608551067897 + }, + "n": 10, + "policy_loss": { + "max": -4.3445531147767724e-05, + "mean": -0.0006323951178690334, + "min": -0.0035095140663869984, + "n": 10, + "std": 0.0009944805630664182 + }, + "value_loss": { + "max": 0.2153652273118496, + "mean": 0.08045755904167891, + "min": 0.040869749151170254, + "n": 10, + "std": 0.04654358081243757 + } + }, + "social": { + "approx_kl": { + "max": 0.004598107310812227, + "mean": 0.00303017191217893, + "min": 0.0012853747145520344, + "n": 10, + "std": 0.0010207158969125706 + }, + "chaser_grad_norm": { + "max": 1.5083578824996948, + "mean": 0.7038940638303757, + "min": 0.36728256940841675, + "n": 10, + "std": 0.3619897429634022 + }, + "chaser_new_fields": { + "max": 63.22500228881836, + "mean": 58.885000991821286, + "min": 55.75, + "n": 10, + "std": 2.30868508603447 + }, + "chaser_partner_vision": { + "max": 0.986750066280365, + "mean": 0.9699750483036041, + "min": 0.9477500319480896, + "n": 10, + "std": 0.010192195843222708 + }, + "chaser_return": { + "max": 16.552501678466797, + "mean": 12.937000370025634, + "min": 9.160000801086426, + "n": 10, + "std": 2.1664790907335263 + }, + "collisions_per_episode": { + "max": 14.475000381469727, + "mean": 10.745000123977661, + "min": 7.550000190734863, + "n": 10, + "std": 2.1296654790690113 + }, + "entropy": { + "max": 0.20668550580739975, + "mean": 0.15412018559873103, + "min": 0.11317355558276176, + "n": 10, + "std": 0.027327137994586925 + }, + "explorer_grad_norm": { + "max": 1.6371502578258514, + "mean": 0.9565056964755059, + "min": 0.60902339220047, + "n": 10, + "std": 0.3402408819474271 + }, + "explorer_new_fields": { + "max": 70.80000305175781, + "mean": 66.54500160217285, + "min": 62.5, + "n": 10, + "std": 2.835794556837627 + }, + "explorer_partner_vision": { + "max": 0.986750066280365, + "mean": 0.9699750483036041, + "min": 0.9477500319480896, + "n": 10, + "std": 0.010192195843222708 + }, + "explorer_return": { + "max": 47.42499923706055, + "mean": 41.021250534057614, + "min": 33.775001525878906, + "n": 10, + "std": 5.261294318837594 + }, + "final_distance": { + "max": 1.8629791736602783, + "mean": 1.6734241247177124, + "min": 1.4801658391952515, + "n": 10, + "std": 0.11725492127768533 + }, + "max_param_abs": { + "max": 2.9420251846313477, + "mean": 2.7518649816513063, + "min": 2.4855704307556152, + "n": 10, + "std": 0.15851267210951459 + }, + "n": 10, + "policy_loss": { + "max": -0.002069776682645852, + "mean": -0.004644430192469429, + "min": -0.007046535808569843, + "n": 10, + "std": 0.001406250863722305 + }, + "value_loss": { + "max": 3.7607646584510803, + "mean": 2.8793665051460264, + "min": 1.877354085445404, + "n": 10, + "std": 0.4693238146909516 + } + } + }, + "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..72b4034731c7a119f86a96961c33ff83f926d373 --- /dev/null +++ b/derived/tables/MANIFEST.json @@ -0,0 +1,67 @@ +{ + "experiment_root": "runs/mouse-run-run-2-transformer", + "generated_at": "2026-07-03T01:24:38.539082+00:00", + "output_hashes": { + "evaluations": "8de65701cca4a6d5b3b6ec45d68a8bccf77698027577760bc3ac20c4f70ad50f", + "exclusions": "31ff7680a00ea1dfa23d4aaa86ec7118d8706edf2005365152a3d680f8a314e1", + "rollouts": "c521dba8fba745966a2515f4198f58eec0e2d6e265ccc164f823d5f04b5cdc3b", + "runs": "bc8443607b1eacf8afb9ec05470609ad4c5c39d1f50d4ce78ff520465f3426d6" + }, + "outputs": { + "evaluations": "runs/tables/mouse-run-run-2-transformer/evaluations.jsonl", + "exclusions": "runs/tables/mouse-run-run-2-transformer/exclusions.jsonl", + "rollouts": "runs/tables/mouse-run-run-2-transformer/rollouts.jsonl", + "runs": "runs/tables/mouse-run-run-2-transformer/runs.jsonl" + }, + "panel_path": "runs/tables/mouse-run-run-2-transformer/panels.json", + "provenance": { + "command": "scripts/transform/build_tables.py runs/mouse-run-run-2-transformer", + "created_at": "2026-07-03T01:24:38.539388+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": 12, + "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..1cb3e5e8dbbc400b2cb8ad994f3d0cd8a0b0730b --- /dev/null +++ b/derived/tables/evaluations.jsonl @@ -0,0 +1,40 @@ +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.030178+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.1368327140808105, "chaser_new_fields": 93.50999450683594, "chaser_partner_vision": 0.0, "chaser_return": 8.491996765136719, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.059999942779541, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.11000061035156, "explorer_partner_vision": 0.0, "explorer_return": -0.949999988079071, "explorer_stuck_run_steps": 2.7200000286102295, "final_distance": 4.908422946929932, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 1, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.168256+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.048107147216797, "chaser_new_fields": 33.82999801635742, "chaser_partner_vision": 0.0, "chaser_return": -3.289998769760132, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 3.059999942779541, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 93.15999603271484, "explorer_partner_vision": 0.0, "explorer_return": 88.68999481201172, "explorer_stuck_run_steps": 0.0, "final_distance": 3.929388999938965, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 2, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.333558+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.137263298034668, "chaser_new_fields": 92.8699951171875, "chaser_partner_vision": 0.0, "chaser_return": 8.371996879577637, "chaser_stuck_run_steps": 0.029999999329447746, "collisions_per_episode": 3.3499999046325684, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.189998626708984, "explorer_partner_vision": 0.0, "explorer_return": -0.8899999856948853, "explorer_stuck_run_steps": 2.75, "final_distance": 4.113985061645508, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 3, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.470364+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.043961048126221, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.0, "chaser_return": -3.2999987602233887, "chaser_stuck_run_steps": 2.6399998664855957, "collisions_per_episode": 3.059999942779541, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 93.68999481201172, "explorer_partner_vision": 0.0, "explorer_return": 89.37999725341797, "explorer_stuck_run_steps": 0.04999999701976776, "final_distance": 3.96620774269104, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 4, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.635312+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.096349239349365, "chaser_new_fields": 93.15999603271484, "chaser_partner_vision": 0.0, "chaser_return": 8.377996444702148, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.8999998569488525, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 33.07999801635742, "explorer_partner_vision": 0.0, "explorer_return": -1.2949999570846558, "explorer_stuck_run_steps": 2.740000009536743, "final_distance": 4.303606033325195, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 5, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.771588+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.124117374420166, "chaser_new_fields": 33.93000030517578, "chaser_partner_vision": 0.0, "chaser_return": -3.2579989433288574, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 2.419999837875366, "degenerate_episodes": 90, "degenerate_fraction": 0.9, "episodes": 100, "explorer_new_fields": 95.88999938964844, "explorer_partner_vision": 0.0, "explorer_return": 93.05500030517578, "explorer_stuck_run_steps": 0.0, "final_distance": 3.837200880050659, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 6, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.936980+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.091452598571777, "chaser_new_fields": 92.63999938964844, "chaser_partner_vision": 0.0, "chaser_return": 8.311995506286621, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 3.3299999237060547, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.11000061035156, "explorer_partner_vision": 0.0, "explorer_return": -1.3249999284744263, "explorer_stuck_run_steps": 2.6499998569488525, "final_distance": 4.107217311859131, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 7, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.073269+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.068935871124268, "chaser_new_fields": 34.20000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.2079992294311523, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 2.7200000286102295, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 94.29999542236328, "explorer_partner_vision": 0.0, "explorer_return": 90.45999908447266, "explorer_stuck_run_steps": 0.0, "final_distance": 4.635434627532959, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 8, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.241084+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.109180927276611, "chaser_new_fields": 92.5199966430664, "chaser_partner_vision": 0.0, "chaser_return": 8.313996315002441, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 2.859999895095825, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 33.06999969482422, "explorer_partner_vision": 0.0, "explorer_return": -0.9799999594688416, "explorer_stuck_run_steps": 2.7200000286102295, "final_distance": 4.262806415557861, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 9, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.377530+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.085238456726074, "chaser_new_fields": 34.36000061035156, "chaser_partner_vision": 0.0, "chaser_return": -3.1899986267089844, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 2.9800000190734863, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 93.37999725341797, "explorer_partner_vision": 0.0, "explorer_return": 89.00499725341797, "explorer_stuck_run_steps": 0.0, "final_distance": 4.054072380065918, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 10, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.531673+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.154397487640381, "chaser_new_fields": 93.55999755859375, "chaser_partner_vision": 0.0, "chaser_return": 8.541996955871582, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 2.6599998474121094, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 32.939998626708984, "explorer_partner_vision": 0.0, "explorer_return": -1.1449999809265137, "explorer_stuck_run_steps": 2.759999990463257, "final_distance": 4.512866973876953, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 11, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.668574+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.100773334503174, "chaser_new_fields": 33.84000015258789, "chaser_partner_vision": 0.0, "chaser_return": -3.279998779296875, "chaser_stuck_run_steps": 2.5799999237060547, "collisions_per_episode": 2.5899999141693115, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 96.0, "explorer_partner_vision": 0.0, "explorer_return": 92.94999694824219, "explorer_stuck_run_steps": 0.009999999776482582, "final_distance": 4.033876895904541, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 12, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.832755+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.138570308685303, "chaser_new_fields": 92.82999420166016, "chaser_partner_vision": 0.0, "chaser_return": 8.367996215820312, "chaser_stuck_run_steps": 0.05999999865889549, "collisions_per_episode": 3.359999895095825, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.21999740600586, "explorer_partner_vision": 0.0, "explorer_return": -0.9350000023841858, "explorer_stuck_run_steps": 2.7100000381469727, "final_distance": 4.475497722625732, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 13, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.968950+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.129418849945068, "chaser_new_fields": 33.880001068115234, "chaser_partner_vision": 0.0, "chaser_return": -3.269998788833618, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 2.319999933242798, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 96.38999938964844, "explorer_partner_vision": 0.0, "explorer_return": 93.50499725341797, "explorer_stuck_run_steps": 0.0, "final_distance": 3.9629032611846924, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 14, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.133560+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.0569047927856445, "chaser_new_fields": 90.43000030517578, "chaser_partner_vision": 0.0, "chaser_return": 7.805994510650635, "chaser_stuck_run_steps": 0.05999999865889549, "collisions_per_episode": 4.210000038146973, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 33.05999755859375, "explorer_partner_vision": 0.0, "explorer_return": -1.3399999141693115, "explorer_stuck_run_steps": 2.690000057220459, "final_distance": 4.142829418182373, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 15, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.270325+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.080101013183594, "chaser_new_fields": 34.290000915527344, "chaser_partner_vision": 0.0, "chaser_return": -3.1859986782073975, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 2.919999837875366, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 93.5199966430664, "explorer_partner_vision": 0.0, "explorer_return": 89.45499420166016, "explorer_stuck_run_steps": 0.009999999776482582, "final_distance": 4.23733377456665, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 16, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.435106+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.084146022796631, "chaser_new_fields": 93.05999755859375, "chaser_partner_vision": 0.0, "chaser_return": 8.329996109008789, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 3.9700000286102295, "degenerate_episodes": 82, "degenerate_fraction": 0.82, "episodes": 100, "explorer_new_fields": 32.97999954223633, "explorer_partner_vision": 0.0, "explorer_return": -1.59499990940094, "explorer_stuck_run_steps": 2.6399998664855957, "final_distance": 4.020629405975342, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 17, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.571327+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.054930686950684, "chaser_new_fields": 33.72999954223633, "chaser_partner_vision": 0.0, "chaser_return": -3.3139989376068115, "chaser_stuck_run_steps": 2.690000057220459, "collisions_per_episode": 2.950000047683716, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 94.97000122070312, "explorer_partner_vision": 0.0, "explorer_return": 91.67499542236328, "explorer_stuck_run_steps": 0.0, "final_distance": 3.9026472568511963, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 18, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.744621+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.109021186828613, "chaser_new_fields": 93.91999816894531, "chaser_partner_vision": 0.0, "chaser_return": 8.579997062683105, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.25, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.209999084472656, "explorer_partner_vision": 0.0, "explorer_return": -1.0399999618530273, "explorer_stuck_run_steps": 2.7100000381469727, "final_distance": 5.0355939865112305, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 19, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.881030+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.061388969421387, "chaser_new_fields": 34.25, "chaser_partner_vision": 0.0, "chaser_return": -3.2059991359710693, "chaser_stuck_run_steps": 2.6399998664855957, "collisions_per_episode": 2.950000047683716, "degenerate_episodes": 90, "degenerate_fraction": 0.9, "episodes": 100, "explorer_new_fields": 92.62999725341797, "explorer_partner_vision": 0.0, "explorer_return": 87.93999481201172, "explorer_stuck_run_steps": 0.08999999612569809, "final_distance": 4.295379638671875, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 20, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.053956+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5451525449752808, "chaser_new_fields": 26.189998626708984, "chaser_partner_vision": 0.9635999798774719, "chaser_return": 42.83700180053711, "chaser_stuck_run_steps": 0.07999999821186066, "collisions_per_episode": 43.98999786376953, "degenerate_episodes": 99, "degenerate_fraction": 0.99, "episodes": 100, "explorer_new_fields": 23.189998626708984, "explorer_partner_vision": 0.9635999798774719, "explorer_return": -44.709999084472656, "explorer_stuck_run_steps": 3.7799999713897705, "final_distance": 1.3304944038391113, "same_state_run_steps": 0.04999999701976776}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 21, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.198234+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.476309776306152, "chaser_new_fields": 34.02000045776367, "chaser_partner_vision": 0.17139999568462372, "chaser_return": -2.9389989376068115, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.25, "degenerate_episodes": 95, "degenerate_fraction": 0.95, "episodes": 100, "explorer_new_fields": 30.420000076293945, "explorer_partner_vision": 0.17139999568462372, "explorer_return": -4.630000114440918, "explorer_stuck_run_steps": 3.75, "final_distance": 6.77753210067749, "same_state_run_steps": 0.2800000011920929}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "schema_version": 3, "seed": 0, "source_line": 22, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.371819+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.588372826576233, "chaser_new_fields": 27.989999771118164, "chaser_partner_vision": 0.9638999700546265, "chaser_return": 36.32500076293945, "chaser_stuck_run_steps": 0.3499999940395355, "collisions_per_episode": 37.72999954223633, "degenerate_episodes": 96, "degenerate_fraction": 0.96, "episodes": 100, "explorer_new_fields": 23.729999542236328, "explorer_partner_vision": 0.9638999700546265, "explorer_return": -40.91999816894531, "explorer_stuck_run_steps": 3.4099998474121094, "final_distance": 1.5478826761245728, "same_state_run_steps": 0.07999999821186066}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 23, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.516274+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.472661018371582, "chaser_new_fields": 33.97999954223633, "chaser_partner_vision": 0.17270000278949738, "chaser_return": -2.8129992485046387, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.3700000047683716, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 36.34000015258789, "explorer_partner_vision": 0.17270000278949738, "explorer_return": 4.159999847412109, "explorer_stuck_run_steps": 6.259999752044678, "final_distance": 6.965867042541504, "same_state_run_steps": 0.6599999666213989}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "schema_version": 3, "seed": 0, "source_line": 24, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.695114+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5360196828842163, "chaser_new_fields": 26.75, "chaser_partner_vision": 0.9630999565124512, "chaser_return": 44.46000289916992, "chaser_stuck_run_steps": 0.3499999940395355, "collisions_per_episode": 45.36000061035156, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 23.31999969482422, "explorer_partner_vision": 0.9630999565124512, "explorer_return": -45.91999816894531, "explorer_stuck_run_steps": 3.6499998569488525, "final_distance": 1.4446792602539062, "same_state_run_steps": 0.07000000029802322}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 25, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.839830+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.161418914794922, "chaser_new_fields": 34.11000061035156, "chaser_partner_vision": 0.19539998471736908, "chaser_return": -2.780999183654785, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.3700000047683716, "degenerate_episodes": 93, "degenerate_fraction": 0.93, "episodes": 100, "explorer_new_fields": 43.939998626708984, "explorer_partner_vision": 0.19539998471736908, "explorer_return": 15.484999656677246, "explorer_stuck_run_steps": 3.3299999237060547, "final_distance": 6.8849663734436035, "same_state_run_steps": 0.26999998092651367}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "schema_version": 3, "seed": 0, "source_line": 26, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.012176+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.4947947263717651, "chaser_new_fields": 26.31999969482422, "chaser_partner_vision": 0.9711999893188477, "chaser_return": 43.392005920410156, "chaser_stuck_run_steps": 0.12999999523162842, "collisions_per_episode": 44.439998626708984, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 22.15999984741211, "explorer_partner_vision": 0.9711999893188477, "explorer_return": -46.93000030517578, "explorer_stuck_run_steps": 3.7200000286102295, "final_distance": 1.3213845491409302, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 27, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.157169+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.446540355682373, "chaser_new_fields": 34.09000015258789, "chaser_partner_vision": 0.17020000517368317, "chaser_return": -2.8699989318847656, "chaser_stuck_run_steps": 2.6399998664855957, "collisions_per_episode": 0.29999998211860657, "degenerate_episodes": 98, "degenerate_fraction": 0.98, "episodes": 100, "explorer_new_fields": 46.5, "explorer_partner_vision": 0.17020000517368317, "explorer_return": 19.404998779296875, "explorer_stuck_run_steps": 6.899999618530273, "final_distance": 6.924410343170166, "same_state_run_steps": 0.5199999809265137}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "schema_version": 3, "seed": 0, "source_line": 28, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.315288+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5326578617095947, "chaser_new_fields": 25.40999984741211, "chaser_partner_vision": 0.961899995803833, "chaser_return": 42.152000427246094, "chaser_stuck_run_steps": 0.14000000059604645, "collisions_per_episode": 43.459999084472656, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 22.09000015258789, "explorer_partner_vision": 0.961899995803833, "explorer_return": -45.70499801635742, "explorer_stuck_run_steps": 3.9800000190734863, "final_distance": 1.2944831848144531, "same_state_run_steps": 0.07000000029802322}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 29, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.459441+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.819681167602539, "chaser_new_fields": 34.09000015258789, "chaser_partner_vision": 0.13769999146461487, "chaser_return": -2.916998863220215, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.25, "degenerate_episodes": 95, "degenerate_fraction": 0.95, "episodes": 100, "explorer_new_fields": 36.32999801635742, "explorer_partner_vision": 0.13769999146461487, "explorer_return": 4.190000057220459, "explorer_stuck_run_steps": 7.349999904632568, "final_distance": 6.673317909240723, "same_state_run_steps": 0.699999988079071}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "schema_version": 3, "seed": 0, "source_line": 30, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.617311+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.6489869356155396, "chaser_new_fields": 28.6299991607666, "chaser_partner_vision": 0.9557999968528748, "chaser_return": 36.21500015258789, "chaser_stuck_run_steps": 0.19999998807907104, "collisions_per_episode": 37.46999740600586, "degenerate_episodes": 96, "degenerate_fraction": 0.96, "episodes": 100, "explorer_new_fields": 24.939998626708984, "explorer_partner_vision": 0.9557999968528748, "explorer_return": -38.09000015258789, "explorer_stuck_run_steps": 3.2899999618530273, "final_distance": 1.4506957530975342, "same_state_run_steps": 0.05999999865889549}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 31, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.761441+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.512670040130615, "chaser_new_fields": 33.98999786376953, "chaser_partner_vision": 0.15479999780654907, "chaser_return": -2.978998899459839, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.20999999344348907, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 41.18000030517578, "explorer_partner_vision": 0.15479999780654907, "explorer_return": 11.529999732971191, "explorer_stuck_run_steps": 3.419999837875366, "final_distance": 6.647017478942871, "same_state_run_steps": 0.38999998569488525}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "schema_version": 3, "seed": 0, "source_line": 32, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.918824+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.667794942855835, "chaser_new_fields": 27.760000228881836, "chaser_partner_vision": 0.9440000057220459, "chaser_return": 36.88300323486328, "chaser_stuck_run_steps": 0.8299999833106995, "collisions_per_episode": 38.290000915527344, "degenerate_episodes": 98, "degenerate_fraction": 0.98, "episodes": 100, "explorer_new_fields": 25.18000030517578, "explorer_partner_vision": 0.9440000057220459, "explorer_return": -38.994998931884766, "explorer_stuck_run_steps": 3.5999999046325684, "final_distance": 1.4672372341156006, "same_state_run_steps": 0.35999998450279236}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 33, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.062945+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.227531433105469, "chaser_new_fields": 34.02000045776367, "chaser_partner_vision": 0.20059996843338013, "chaser_return": -2.903998851776123, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.2800000011920929, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 41.709999084472656, "explorer_partner_vision": 0.20059996843338013, "explorer_return": 12.244999885559082, "explorer_stuck_run_steps": 5.0, "final_distance": 6.640198707580566, "same_state_run_steps": 0.38999998569488525}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "schema_version": 3, "seed": 0, "source_line": 34, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.220422+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5710211992263794, "chaser_new_fields": 25.19999885559082, "chaser_partner_vision": 0.9613999724388123, "chaser_return": 40.952999114990234, "chaser_stuck_run_steps": 0.19999998807907104, "collisions_per_episode": 42.369998931884766, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 21.5, "explorer_partner_vision": 0.9613999724388123, "explorer_return": -46.85499954223633, "explorer_stuck_run_steps": 3.7200000286102295, "final_distance": 1.3192702531814575, "same_state_run_steps": 0.05999999865889549}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 35, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.364863+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.704348087310791, "chaser_new_fields": 33.939998626708984, "chaser_partner_vision": 0.15079998970031738, "chaser_return": -2.963999032974243, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.23999999463558197, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 30.35999870300293, "explorer_partner_vision": 0.15079998970031738, "explorer_return": -4.670000076293945, "explorer_stuck_run_steps": 15.00999927520752, "final_distance": 6.317955017089844, "same_state_run_steps": 1.3399999141693115}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "schema_version": 3, "seed": 0, "source_line": 36, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.522956+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5945370197296143, "chaser_new_fields": 26.549999237060547, "chaser_partner_vision": 0.9594999551773071, "chaser_return": 34.46500015258789, "chaser_stuck_run_steps": 0.08999999612569809, "collisions_per_episode": 36.189998626708984, "degenerate_episodes": 99, "degenerate_fraction": 0.99, "episodes": 100, "explorer_new_fields": 22.969999313354492, "explorer_partner_vision": 0.9594999551773071, "explorer_return": -40.5099983215332, "explorer_stuck_run_steps": 3.7200000286102295, "final_distance": 1.4907904863357544, "same_state_run_steps": 0.03999999910593033}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 37, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.667220+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.724240779876709, "chaser_new_fields": 33.96999740600586, "chaser_partner_vision": 0.13729999959468842, "chaser_return": -2.9429991245269775, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 0.25, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 34.59000015258789, "explorer_partner_vision": 0.13729999959468842, "explorer_return": 1.5499999523162842, "explorer_stuck_run_steps": 14.15999984741211, "final_distance": 6.654120445251465, "same_state_run_steps": 1.0299999713897705}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "schema_version": 3, "seed": 0, "source_line": 38, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.825633+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5335965156555176, "chaser_new_fields": 27.619998931884766, "chaser_partner_vision": 0.964900016784668, "chaser_return": 42.8380012512207, "chaser_stuck_run_steps": 0.35999998450279236, "collisions_per_episode": 43.779998779296875, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 24.18000030517578, "explorer_partner_vision": 0.964900016784668, "explorer_return": -44.54499816894531, "explorer_stuck_run_steps": 3.5199999809265137, "final_distance": 1.3675272464752197, "same_state_run_steps": 0.08999999612569809}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 39, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.970033+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.735382080078125, "chaser_new_fields": 34.09000015258789, "chaser_partner_vision": 0.14399999380111694, "chaser_return": -2.9249989986419678, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.25, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 34.63999938964844, "explorer_partner_vision": 0.14399999380111694, "explorer_return": 1.6999999284744263, "explorer_stuck_run_steps": 9.229999542236328, "final_distance": 6.582851409912109, "same_state_run_steps": 0.5699999928474426}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "schema_version": 3, "seed": 0, "source_line": 40, "source_path": "runs/mouse-run-run-2-transformer/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..e0fb9b7b38067056781808a42145db92bc9ef677 --- /dev/null +++ b/derived/tables/exclusions.jsonl @@ -0,0 +1,12 @@ +{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 13, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 10, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 2, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 11, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 5, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 4, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 5, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 5, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors", "threshold_fraction": 0.01, "total_count": 25} +{"excluded_count": 1, "reason": "paper_degenerate_episode_v1", "schema_version": 1, "source_rollout": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors", "threshold_fraction": 0.01, "total_count": 25} 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..33335673bf07cb328bfff52ea8896a4060f86799 --- /dev/null +++ b/derived/tables/rollouts.jsonl @@ -0,0 +1,20 @@ +{"attempt_id": "attempt_01", "batch_size": 25, "checkpoint": "runs/mouse-run-run-2-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:19.731502+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "source_line": 1, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:20.616845+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "source_line": 2, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:21.500858+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "source_line": 3, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:22.386345+00:00", "degenerate_episode_count": 13, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "source_line": 4, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:23.279182+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "source_line": 5, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:24.164102+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "source_line": 6, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:25.051469+00:00", "degenerate_episode_count": 10, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "source_line": 7, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:25.937613+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "source_line": 8, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:26.826889+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "source_line": 9, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:27.714387+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "source_line": 10, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:28.637667+00:00", "degenerate_episode_count": 2, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "schema_version": 2, "seed": 0, "source_line": 11, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:29.559684+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "schema_version": 2, "seed": 0, "source_line": 12, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:30.484944+00:00", "degenerate_episode_count": 11, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "schema_version": 2, "seed": 0, "source_line": 13, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:31.410494+00:00", "degenerate_episode_count": 5, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "schema_version": 2, "seed": 0, "source_line": 14, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:32.334168+00:00", "degenerate_episode_count": 4, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "schema_version": 2, "seed": 0, "source_line": 15, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:33.266584+00:00", "degenerate_episode_count": 5, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "schema_version": 2, "seed": 0, "source_line": 16, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:34.189914+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-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "schema_version": 2, "seed": 0, "source_line": 17, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:35.152778+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "schema_version": 2, "seed": 0, "source_line": 18, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:36.085812+00:00", "degenerate_episode_count": 5, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "schema_version": 2, "seed": 0, "source_line": 19, "source_path": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:37.008624+00:00", "degenerate_episode_count": 1, "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episode_count": 25, "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "schema_version": 2, "seed": 0, "source_line": 20, "source_path": "runs/mouse-run-run-2-transformer/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..c71740c0823b83692d02e0dc2d494fd4be33278b --- /dev/null +++ b/derived/tables/runs.jsonl @@ -0,0 +1,20 @@ +{"attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19540.62696071295, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19646.28115725203, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19551.709339142893, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19539.363500279956, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19559.85987384396, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19523.735058769933, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19464.16788898199, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19529.17995151691, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19471.327225218993, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19415.301637928, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19670.30402108794, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19568.46814744, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19568.42091672402, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19557.178920289967, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19601.35430799797, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19533.043117783964, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19480.819718190003, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19530.44018116698, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19499.830016753986, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "run_dir": "runs/mouse-run-run-2-transformer/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-transformer/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": 266}, "checkpoint_ok": true, "config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 19455.73302768101, "environment_steps": 80000000, "error": null, "eta_seconds": 0.0, "experiment": "mouse-run-run-2-transformer", "healthy": true, "is_latest_successful": true, "metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "run_dir": "runs/mouse-run-run-2-transformer/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/launch_gate.json b/launch_gate.json new file mode 100644 index 0000000000000000000000000000000000000000..f7b3507bb7ea83f8d3dda387bbc2627b3b31fee4 --- /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-transformer/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_transformer.json", + "config_sha256": "2bbd0502a075bf0d5c49b248539f3759048c41a1b20e63efbe677e891decf7b1", + "created_at": "2026-07-02T14:31:49.723425+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_transformer.json --triton-equivalence runs/mouse-run-run-2-transformer/triton_equivalence.json", + "created_at": "2026-07-02T14:31:49.723576+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-transformer/triton_equivalence.json" +} diff --git a/manifest.json b/manifest.json new file mode 100644 index 0000000000000000000000000000000000000000..6bdf27d541e7169d3822ca2f825f9bdd692ba9a7 --- /dev/null +++ b/manifest.json @@ -0,0 +1,423 @@ +{ + "config": { + "architecture": "transformer", + "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": "6e419bd117b1aae2fbbef85bee470cef165112bab56d10a81e5dfbdc5bb6f9c5", + "created_at": "2026-07-02T14:31:50.715150+00:00", + "experiment": "mouse-run-run-2-transformer", + "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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0000_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0000_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0001_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0001_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0002_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0002_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0003_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0003_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0004_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0004_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0005_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0005_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0006_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0006_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0007_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0007_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0008_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0008_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0009_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0009_attempt_01.log", + "resume_from": null, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-transformer.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_transformer.json", + "created_at": "2026-07-02T14:31:50.715276+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..9b1767ef7f6f8f003e941295bdf84e6acd73d9be --- /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:bebcbee7c71f93593b985b9403dddac3f4b9d301088feea80adec365e5c839d3 +size 29760368 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..6f22a010d878f7351f90927ee15e28fc8d7cdad9 --- /dev/null +++ b/non_social/seed_0000/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..a7ee23530a985c8b57c5ea14d6eabdc91f5c67ae --- /dev/null +++ b/non_social/seed_0000/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007858585283328365, "elapsed_seconds": 2.8290907019982114, "env_steps_per_second": 1413.8818515697517, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 15.716384708128398, "eta_seconds": 56578.98494926223, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.003984742708620814, "chaser_grad_norm": 6.097266681492329, "chaser_new_fields": 34.67499923706055, "chaser_partner_vision": 0.0, "chaser_return": -3.130000352859497, "chaser_subspace_norm": 0.03538785129785538, "collisions_per_episode": 2.1500000953674316, "entropy": 2.757209300994873, "explorer_grad_norm": 6.22088485956192, "explorer_new_fields": 34.04999923706055, "explorer_partner_vision": 0.0, "explorer_return": 0.512499988079071, "explorer_subspace_norm": 0.0416051484644413, "final_distance": 5.596360206604004, "max_param_abs": 1.001202940940857, "policy_loss": -0.015378041312104074, "value_loss": 7.977049231529236}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.904332+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.9043489, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13819628580276105, "elapsed_seconds": 497.50662888993975, "env_steps_per_second": 4020.0469377915515, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.389655146307681, "eta_seconds": 19402.75852670765, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005213568198339402, "chaser_grad_norm": 0.6902080327272415, "chaser_new_fields": 85.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.1100006103515625, "chaser_subspace_norm": 12.22103214263916, "collisions_per_episode": 1.399999976158142, "entropy": 0.10881881602108479, "explorer_grad_norm": 4.12031877040863, "explorer_new_fields": 89.875, "explorer_partner_vision": 0.0, "explorer_return": 84.4375, "explorer_subspace_norm": 15.985239028930664, "final_distance": 3.488882541656494, "max_param_abs": 1.2841757535934448, "policy_loss": -0.003982506634656602, "value_loss": 0.554658830165863}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:09.581869+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003209.5818808, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27406795207886414, "elapsed_seconds": 986.6446274839109, "env_steps_per_second": 4054.144611520958, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.207291089498418, "eta_seconds": 18746.247922194307, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0009871974141866247, "chaser_grad_norm": 0.08980551920831203, "chaser_new_fields": 88.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.755000591278076, "chaser_subspace_norm": 17.67871856689453, "collisions_per_episode": 0.7750000357627869, "entropy": 0.03490936104208231, "explorer_grad_norm": 10.575141370296478, "explorer_new_fields": 93.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 89.3499984741211, "explorer_subspace_norm": 29.749900817871094, "final_distance": 3.751117706298828, "max_param_abs": 1.4041469097137451, "policy_loss": -0.002187443812843526, "value_loss": 0.4212941825389862}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:18.719867+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003698.719881, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4078779447350017, "elapsed_seconds": 1468.3606010460062, "env_steps_per_second": 4086.1897245988625, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.030494651731687, "eta_seconds": 18109.780746234075, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0010331508172161108, "chaser_grad_norm": 0.11727817915380001, "chaser_new_fields": 90.625, "chaser_partner_vision": 0.0, "chaser_return": 8.120000839233398, "chaser_subspace_norm": 22.310590744018555, "collisions_per_episode": 0.22500000894069672, "entropy": 0.01846518414095044, "explorer_grad_norm": 3.1013119220733643, "explorer_new_fields": 94.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.26250457763672, "explorer_subspace_norm": 40.93424606323242, "final_distance": 3.4480628967285156, "max_param_abs": 1.505894422531128, "policy_loss": -0.0012613329393786898, "value_loss": 0.07974300533533096}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:20.435839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004180.43585, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5416774474972044, "elapsed_seconds": 1950.0388109899359, "env_steps_per_second": 4102.482450561487, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.8750970274748395, "eta_seconds": 17550.349298909423, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0010745732343774295, "chaser_grad_norm": 0.32581115886569023, "chaser_new_fields": 87.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.460000514984131, "chaser_subspace_norm": 26.053730010986328, "collisions_per_episode": 1.524999976158142, "entropy": 0.04285650607198477, "explorer_grad_norm": 2.352870851755142, "explorer_new_fields": 92.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 88.11250305175781, "explorer_subspace_norm": 48.09144973754883, "final_distance": 3.754744052886963, "max_param_abs": 1.5702295303344727, "policy_loss": -0.0025823999991416713, "value_loss": 0.5012058317661285}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:22.114052+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004662.1140668, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6754466422619427, "elapsed_seconds": 2431.607912142994, "env_steps_per_second": 4112.50512472092, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.7281264958336, "eta_seconds": 17021.255385000957, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0009930499389909642, "chaser_grad_norm": 0.04422433767467737, "chaser_new_fields": 89.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.875000953674316, "chaser_subspace_norm": 27.880756378173828, "collisions_per_episode": 0.7750000357627869, "entropy": 0.016509887762367725, "explorer_grad_norm": 1.9961952567100525, "explorer_new_fields": 94.5, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 57.431854248046875, "final_distance": 4.026604652404785, "max_param_abs": 1.6253595352172852, "policy_loss": -0.0012972251977743099, "value_loss": 0.20397597178816795}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:23.683150+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005143.6831622, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8092757643613732, "elapsed_seconds": 2913.3927517009433, "env_steps_per_second": 4118.909128538872, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.585895998047781, "eta_seconds": 16509.225592972012, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0027243835385055704, "chaser_grad_norm": 0.4044172838330269, "chaser_new_fields": 90.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.135001182556152, "chaser_subspace_norm": 29.07428550720215, "collisions_per_episode": 0.3499999940395355, "entropy": 0.013051415095105767, "explorer_grad_norm": 0.7505706623196602, "explorer_new_fields": 94.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 68.80549621582031, "final_distance": 3.6912147998809814, "max_param_abs": 1.6739788055419922, "policy_loss": -0.003254898154175123, "value_loss": 0.12792102247476578}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:25.467991+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005625.4680054, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9452245050919332, "elapsed_seconds": 3402.80821833096, "env_steps_per_second": 4114.248909057486, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.4560583811476855, "eta_seconds": 16041.810172131667, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0001970463801899136, "chaser_grad_norm": 0.0663881553336978, "chaser_new_fields": 90.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.095001220703125, "chaser_subspace_norm": 34.47535705566406, "collisions_per_episode": 0.699999988079071, "entropy": 0.009223120054230094, "explorer_grad_norm": 0.439394012093544, "explorer_new_fields": 94.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 71.96785736083984, "final_distance": 4.070150852203369, "max_param_abs": 1.7069555521011353, "policy_loss": -0.0004109265974836518, "value_loss": 0.12886791303753853}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:34.883457+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006114.8834677, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0821224993897098, "elapsed_seconds": 3895.640997802955, "env_steps_per_second": 4107.1546400255065, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.3284899975588385, "eta_seconds": 15582.563991211819, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.002061240742013748, "chaser_grad_norm": 0.14911585487425327, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 36.80433654785156, "collisions_per_episode": 0.800000011920929, "entropy": 0.007869288441725075, "explorer_grad_norm": 0.9927732795476913, "explorer_new_fields": 94.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 77.96184539794922, "final_distance": 3.4226481914520264, "max_param_abs": 1.741771936416626, "policy_loss": -0.003020491153939986, "value_loss": 0.147367924451828}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:36:47.716237+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006607.716251, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2179189361097007, "elapsed_seconds": 4384.508169994922, "env_steps_per_second": 4105.363544121494, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.195054113266747, "eta_seconds": 15102.194807760288, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0016153937390637196, "chaser_grad_norm": 0.06304028397426009, "chaser_new_fields": 93.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 38.633880615234375, "collisions_per_episode": 0.800000011920929, "entropy": 0.00674564519431442, "explorer_grad_norm": 0.7198946475982666, "explorer_new_fields": 94.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.75, "explorer_subspace_norm": 83.59247589111328, "final_distance": 3.769803762435913, "max_param_abs": 1.7817188501358032, "policy_loss": -0.0006622512721223828, "value_loss": 0.11152646876871586}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:44:56.583410+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007096.5834255, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.357942673002221, "elapsed_seconds": 4888.593622807995, "env_steps_per_second": 4091.156177655866, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.073828019006663, "eta_seconds": 14665.780868423986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0002423651793477255, "chaser_grad_norm": 0.038826094940304756, "chaser_new_fields": 92.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.480000495910645, "chaser_subspace_norm": 37.68648910522461, "collisions_per_episode": 0.6000000238418579, "entropy": 0.00865237764082849, "explorer_grad_norm": 0.6159049421548843, "explorer_new_fields": 94.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.7874984741211, "explorer_subspace_norm": 86.39503479003906, "final_distance": 3.6380271911621094, "max_param_abs": 1.810005784034729, "policy_loss": -0.0006053061515558511, "value_loss": 0.08581102639436722}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:20.668862+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007600.6688757, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.496551224394158, "elapsed_seconds": 5387.584407818969, "env_steps_per_second": 4083.4627051172565, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9454532279482346, "eta_seconds": 14203.631620613645, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0007384293112006057, "chaser_grad_norm": 0.12568512931466103, "chaser_new_fields": 92.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.55000114440918, "chaser_subspace_norm": 41.24983215332031, "collisions_per_episode": 0.45000001788139343, "entropy": 0.008910412434488535, "explorer_grad_norm": 0.9198387712240219, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.2750015258789, "explorer_subspace_norm": 90.77873992919922, "final_distance": 3.517383575439453, "max_param_abs": 1.8441685438156128, "policy_loss": -0.0011138532813639301, "value_loss": 0.06246422231197357}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:39.659647+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008099.6596582, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6330133805391638, "elapsed_seconds": 5878.8481699409895, "env_steps_per_second": 4082.4323585551806, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.8103645545913816, "eta_seconds": 13717.312396528974, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0002877887917747124, "chaser_grad_norm": 0.0709311356768012, "chaser_new_fields": 92.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.465001106262207, "chaser_subspace_norm": 39.979000091552734, "collisions_per_episode": 0.75, "entropy": 0.006909179617650807, "explorer_grad_norm": 0.5209658145904541, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.67500305175781, "explorer_subspace_norm": 97.17213439941406, "final_distance": 3.5823585987091064, "max_param_abs": 1.865701675415039, "policy_loss": -0.0007798157297145281, "value_loss": 0.11099696718156338}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:50.923411+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008590.9234254, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.767157393639144, "elapsed_seconds": 6361.766617100919, "env_steps_per_second": 4086.915092123939, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6702499714043766, "eta_seconds": 13212.899897055755, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0007893479407814347, "chaser_grad_norm": 0.04300319030880928, "chaser_new_fields": 92.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.510001182556152, "chaser_subspace_norm": 45.24545669555664, "collisions_per_episode": 0.625, "entropy": 0.005201557534746826, "explorer_grad_norm": 0.36915695667266846, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.01250457763672, "explorer_subspace_norm": 100.9892349243164, "final_distance": 3.551710605621338, "max_param_abs": 1.8893122673034668, "policy_loss": -0.00030515577702328045, "value_loss": 0.10670126788318157}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:53.841855+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009073.8418665, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.9058069222163694, "elapsed_seconds": 6860.90491997893, "env_steps_per_second": 4081.094305572448, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5393557126875432, "eta_seconds": 12741.680565675155, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0005970458028607961, "chaser_grad_norm": 0.08210457861423492, "chaser_new_fields": 92.625, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 44.65858459472656, "collisions_per_episode": 0.5, "entropy": 0.006243823445402086, "explorer_grad_norm": 0.2898806743323803, "explorer_new_fields": 95.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 101.96656036376953, "final_distance": 3.8500595092773438, "max_param_abs": 1.9082915782928467, "policy_loss": -0.0005075841061774611, "value_loss": 0.06872157379984856}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:26:12.980159+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009572.9801733, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0401156900350017, "elapsed_seconds": 7344.416484126006, "env_steps_per_second": 4084.735671627701, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.400192816725003, "eta_seconds": 12240.694140210011, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0012023549742483652, "chaser_grad_norm": 0.18835537135601044, "chaser_new_fields": 92.875, "chaser_partner_vision": 0.0, "chaser_return": 8.51500129699707, "chaser_subspace_norm": 47.02663040161133, "collisions_per_episode": 0.875, "entropy": 0.004316961974836886, "explorer_grad_norm": 0.742193266749382, "explorer_new_fields": 94.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 106.36807250976562, "final_distance": 3.8010079860687256, "max_param_abs": 1.9282002449035645, "policy_loss": -0.0011232557701852386, "value_loss": 0.10672311298549175}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:34:16.491724+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010056.491738, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.178467757531649, "elapsed_seconds": 7842.483927113935, "env_steps_per_second": 4080.33988942788, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.267701636297473, "eta_seconds": 11763.725890670903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": -1.6572943833947917e-05, "chaser_grad_norm": 0.03920569969341159, "chaser_new_fields": 93.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 47.99747085571289, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004179352545179427, "explorer_grad_norm": 0.42505457252264023, "explorer_new_fields": 94.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.5250015258789, "explorer_subspace_norm": 108.12300872802734, "final_distance": 3.975580930709839, "max_param_abs": 1.9446109533309937, "policy_loss": -8.778000239839656e-05, "value_loss": 0.11882751807570457}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:34.559169+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010554.5591846, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3151148458847053, "elapsed_seconds": 8334.413445184939, "env_steps_per_second": 4079.4712457710993, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1322142032557774, "eta_seconds": 11275.971131720798, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00039934206656147553, "chaser_grad_norm": 0.054306747391819954, "chaser_new_fields": 93.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 44.896263122558594, "collisions_per_episode": 0.675000011920929, "entropy": 0.006878925370983779, "explorer_grad_norm": 0.43170301616191864, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 110.83919525146484, "final_distance": 4.0531816482543945, "max_param_abs": 1.9534330368041992, "policy_loss": -0.0007233839633471462, "value_loss": 0.06848301365971565}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:46.488683+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011046.4886954, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.449597211649169, "elapsed_seconds": 8818.54996193701, "env_steps_per_second": 4082.3037977201116, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.993952147571207, "eta_seconds": 10778.227731256346, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00018719192692023823, "chaser_grad_norm": 0.11254973709583282, "chaser_new_fields": 93.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.61500072479248, "chaser_subspace_norm": 48.95949935913086, "collisions_per_episode": 0.875, "entropy": 0.0051835766062140465, "explorer_grad_norm": 0.39654700085520744, "explorer_new_fields": 95.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.1624984741211, "explorer_subspace_norm": 114.00794982910156, "final_distance": 4.262425899505615, "max_param_abs": 1.9768918752670288, "policy_loss": -0.0004432122928577087, "value_loss": 0.06671177595853806}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:50.625202+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011530.6252155, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5876557905505373, "elapsed_seconds": 9315.560845981934, "env_steps_per_second": 4079.196156653357, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.860040610608489, "eta_seconds": 10296.14619819056, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005332294335725907, "chaser_grad_norm": 0.0458578048273921, "chaser_new_fields": 93.125, "chaser_partner_vision": 0.0, "chaser_return": 8.59000015258789, "chaser_subspace_norm": 50.698280334472656, "collisions_per_episode": 0.550000011920929, "entropy": 0.00311667169444263, "explorer_grad_norm": 0.554467223584652, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.4625015258789, "explorer_subspace_norm": 115.92314910888672, "final_distance": 4.192587375640869, "max_param_abs": 1.9959673881530762, "policy_loss": -0.00047228195774096093, "value_loss": 0.05310062225908041}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:07:07.636087+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012027.6361022, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7234276721347124, "elapsed_seconds": 9804.339619684964, "env_steps_per_second": 4079.826031290141, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7234276721347124, "eta_seconds": 9804.339619684964, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005473158898638975, "chaser_grad_norm": 0.12831407506018877, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.710000991821289, "chaser_subspace_norm": 52.909427642822266, "collisions_per_episode": 0.699999988079071, "entropy": 0.0029723012703470886, "explorer_grad_norm": 0.2813072130084038, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 116.91361236572266, "final_distance": 4.026191234588623, "max_param_abs": 2.006232500076294, "policy_loss": -0.000196251879156506, "value_loss": 0.07602320238947868}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:15:16.414861+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012516.4148755, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8574518892694485, "elapsed_seconds": 10286.826801370014, "env_steps_per_second": 4082.8917226842373, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.58531361410093, "eta_seconds": 9307.129010763347, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0009713084957361506, "chaser_grad_norm": 0.027999839279800653, "chaser_new_fields": 93.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 54.712059020996094, "collisions_per_episode": 0.75, "entropy": 0.005503859603777528, "explorer_grad_norm": 0.2878112308681011, "explorer_new_fields": 94.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 120.97154235839844, "final_distance": 3.91410756111145, "max_param_abs": 2.0216140747070312, "policy_loss": -0.0005653448201483613, "value_loss": 0.1004257034510374}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:18.902040+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012998.902052, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9921132260405447, "elapsed_seconds": 10771.60761374596, "env_steps_per_second": 4084.812738987106, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4480926394877183, "eta_seconds": 8813.133502155786, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 1.541443885122451e-05, "chaser_grad_norm": 0.029784617945551872, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 53.42091751098633, "collisions_per_episode": 0.625, "entropy": 0.0029498518561013043, "explorer_grad_norm": 0.29834797978401184, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 121.83349609375, "final_distance": 3.73388671875, "max_param_abs": 2.040980339050293, "policy_loss": -9.137296861494448e-05, "value_loss": 0.066783607006073}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:23.682854+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013483.6828663, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1278307616852623, "elapsed_seconds": 11260.190742066945, "env_steps_per_second": 4085.188346601324, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3118749108108463, "eta_seconds": 8322.749678919046, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00011620879817808571, "chaser_grad_norm": 0.09901169408112764, "chaser_new_fields": 93.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 54.40679168701172, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004110373207367957, "explorer_grad_norm": 0.2696717716753483, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 123.82683563232422, "final_distance": 3.829066514968872, "max_param_abs": 2.0539538860321045, "policy_loss": -0.000206822176653354, "value_loss": 0.07687818817794323}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:32.265981+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013972.2659926, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.262044489343049, "elapsed_seconds": 11743.360161634977, "env_steps_per_second": 4087.41615170876, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1746963262286996, "eta_seconds": 7828.906774423318, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00032187812307515884, "chaser_grad_norm": 0.04246761417016387, "chaser_new_fields": 93.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 56.21242141723633, "collisions_per_episode": 0.75, "entropy": 0.003060928371269256, "explorer_grad_norm": 0.5045444145798683, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 125.90714263916016, "final_distance": 3.640484571456909, "max_param_abs": 2.057662010192871, "policy_loss": -0.00028770001560118885, "value_loss": 0.07355882227420807}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:35.435403+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014455.4354193, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.396113327720539, "elapsed_seconds": 12226.00797979394, "env_steps_per_second": 4089.6423495416952, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0376679966323232, "eta_seconds": 7335.604787876364, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0006283227894127719, "chaser_grad_norm": 0.1221990492194891, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.644999504089355, "chaser_subspace_norm": 56.58079528808594, "collisions_per_episode": 0.7750000357627869, "entropy": 0.005023272125981748, "explorer_grad_norm": 0.3120904453098774, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 127.02792358398438, "final_distance": 4.106462001800537, "max_param_abs": 2.0744540691375732, "policy_loss": -0.0007666714633818295, "value_loss": 0.0993267111480236}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:38.083219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014938.08323, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5320635015466704, "elapsed_seconds": 12715.428605568013, "env_steps_per_second": 4089.5200321623056, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9018803469866685, "eta_seconds": 6846.7692491520065, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00033326833698955893, "chaser_grad_norm": 0.04215310048311949, "chaser_new_fields": 93.625, "chaser_partner_vision": 0.0, "chaser_return": 8.710000038146973, "chaser_subspace_norm": 56.401405334472656, "collisions_per_episode": 0.625, "entropy": 0.005254154675640166, "explorer_grad_norm": 0.2833205871284008, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 127.3544921875, "final_distance": 4.4870429039001465, "max_param_abs": 2.089384078979492, "policy_loss": -0.0005055732870147267, "value_loss": 0.09795444644987583}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:47.503846+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015427.5038593, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.668322203079426, "elapsed_seconds": 13205.959931085934, "env_steps_per_second": 4089.0628384300685, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.766229208890094, "eta_seconds": 6358.425152004338, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00034709783561170937, "chaser_grad_norm": 0.0546520808711648, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 56.26631546020508, "collisions_per_episode": 0.625, "entropy": 0.004097803262993693, "explorer_grad_norm": 0.46600425988435745, "explorer_new_fields": 94.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 132.22207641601562, "final_distance": 3.958134412765503, "max_param_abs": 2.109316825866699, "policy_loss": -0.000491810953768379, "value_loss": 0.06270698457956314}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:58.035172+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015918.0351858, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.805968612701658, "elapsed_seconds": 13701.487005725969, "env_steps_per_second": 4087.1476195683813, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6311294054435677, "eta_seconds": 5872.065859596843, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.000709932864020546, "chaser_grad_norm": 0.060600987169891596, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 59.46336364746094, "collisions_per_episode": 0.574999988079071, "entropy": 0.004734171321615577, "explorer_grad_norm": 0.20277123898267746, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 138.1077117919922, "final_distance": 3.830503225326538, "max_param_abs": 2.1345181465148926, "policy_loss": -0.0005144484461050425, "value_loss": 0.0775738526135683}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:13.562247+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016413.562261, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.940078420972762, "elapsed_seconds": 14184.282315501943, "env_steps_per_second": 4089.033107907902, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4945125045069099, "eta_seconds": 5380.245016224875, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0003340224873058517, "chaser_grad_norm": 0.07462147623300552, "chaser_new_fields": 93.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 59.58805465698242, "collisions_per_episode": 0.574999988079071, "entropy": 0.005957852001301944, "explorer_grad_norm": 0.32502011209726334, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 137.80844116210938, "final_distance": 3.650078535079956, "max_param_abs": 2.134597063064575, "policy_loss": -0.0004052768719855493, "value_loss": 0.08263413421809673}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:16.357554+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016896.357567, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.074239472052755, "elapsed_seconds": 14667.262099389918, "env_steps_per_second": 4090.7430162099367, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3580798240175849, "eta_seconds": 4889.087366463305, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.000514724589585569, "chaser_grad_norm": 0.053545817732810974, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.710000991821289, "chaser_subspace_norm": 63.33787536621094, "collisions_per_episode": 0.550000011920929, "entropy": 0.004608422867022455, "explorer_grad_norm": 0.30683189257979393, "explorer_new_fields": 95.625, "explorer_partner_vision": 0.0, "explorer_return": 93.32500457763672, "explorer_subspace_norm": 137.96897888183594, "final_distance": 3.8315236568450928, "max_param_abs": 2.139131546020508, "policy_loss": -0.0006109134045004794, "value_loss": 0.056757569313049316}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:19.337340+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017379.337354, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.210373383232218, "elapsed_seconds": 15157.344179635984, "env_steps_per_second": 4090.4263481261783, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2223664660996763, "eta_seconds": 4400.519277958834, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004188151698392151, "chaser_grad_norm": 0.03207044675946236, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.90999984741211, "chaser_subspace_norm": 61.26824188232422, "collisions_per_episode": 0.800000011920929, "entropy": 0.0030996299465186894, "explorer_grad_norm": 0.38353733718395233, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 139.2194061279297, "final_distance": 3.9234397411346436, "max_param_abs": 2.1418116092681885, "policy_loss": -0.00018357145809488884, "value_loss": 0.05587191041558981}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:29.419418+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017869.4194288, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.345209612597763, "elapsed_seconds": 15642.754605351947, "env_steps_per_second": 4091.3510193468933, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0863024031494408, "eta_seconds": 3910.688651337987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00031747748237398454, "chaser_grad_norm": 0.05185403674840927, "chaser_new_fields": 94.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.9399995803833, "chaser_subspace_norm": 64.73976135253906, "collisions_per_episode": 0.7750000357627869, "entropy": 0.004601132357493043, "explorer_grad_norm": 0.21299517154693604, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 139.50474548339844, "final_distance": 3.724943161010742, "max_param_abs": 2.1471619606018066, "policy_loss": -0.000248220444746039, "value_loss": 0.07776057720184326}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:34.829844+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018354.8298604, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.479314453306369, "elapsed_seconds": 16125.532031902927, "env_steps_per_second": 4092.8882141330214, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9501576113074115, "eta_seconds": 3420.5674007066814, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00019849943035726847, "chaser_grad_norm": 0.04664633749052882, "chaser_new_fields": 95.0, "chaser_partner_vision": 0.0, "chaser_return": 8.975000381469727, "chaser_subspace_norm": 68.20988464355469, "collisions_per_episode": 0.7250000238418579, "entropy": 0.002828369732014835, "explorer_grad_norm": 0.5940320566296577, "explorer_new_fields": 95.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 143.01023864746094, "final_distance": 3.92901611328125, "max_param_abs": 2.1550333499908447, "policy_loss": -0.0003368458753683079, "value_loss": 0.0917948242276907}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:37.607270+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018837.6072822, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.61318730738415, "elapsed_seconds": 16607.47430658294, "env_steps_per_second": 4094.541936039358, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8140918777736734, "eta_seconds": 2930.7307599852243, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00021912654095368467, "chaser_grad_norm": 0.05596489645540714, "chaser_new_fields": 95.0, "chaser_partner_vision": 0.0, "chaser_return": 8.960000038146973, "chaser_subspace_norm": 67.79967498779297, "collisions_per_episode": 0.45000001788139343, "entropy": 0.003220201935619116, "explorer_grad_norm": 0.2399616837501526, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.25, "explorer_subspace_norm": 141.00111389160156, "final_distance": 3.815213918685913, "max_param_abs": 2.1613283157348633, "policy_loss": -0.00019624019481145893, "value_loss": 0.04436328634619713}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:39.549546+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019319.5495608, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.74989137156611, "elapsed_seconds": 17099.608937637997, "env_steps_per_second": 4093.660869981816, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6785559102237301, "eta_seconds": 2442.801276805428, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 5.842784732817963e-05, "chaser_grad_norm": 0.02148897247388959, "chaser_new_fields": 95.125, "chaser_partner_vision": 0.0, "chaser_return": 8.970000267028809, "chaser_subspace_norm": 69.08584594726562, "collisions_per_episode": 0.75, "entropy": 0.0033080510911531746, "explorer_grad_norm": 0.2629040814936161, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 141.89207458496094, "final_distance": 3.8556153774261475, "max_param_abs": 2.177971124649048, "policy_loss": -7.839060096737072e-05, "value_loss": 0.08136884681880474}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:16:51.684176+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019811.684187, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.88537475555081, "elapsed_seconds": 17587.349119982915, "env_steps_per_second": 4093.8517515520807, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5428194172834233, "eta_seconds": 1954.1499022203238, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005076670059834323, "chaser_grad_norm": 0.03680429421365261, "chaser_new_fields": 95.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.994999885559082, "chaser_subspace_norm": 62.626258850097656, "collisions_per_episode": 0.574999988079071, "entropy": 0.003358416724950075, "explorer_grad_norm": 0.37850791588425636, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 145.23422241210938, "final_distance": 4.000051975250244, "max_param_abs": 2.179173469543457, "policy_loss": -0.0003006897167652234, "value_loss": 0.05968277156352997}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:24:59.424359+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020299.4243724, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.020150505876939, "elapsed_seconds": 18072.541821156978, "env_steps_per_second": 4094.6094208713043, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4070392302062382, "eta_seconds": 1465.3412287424576, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 1.8009533259876614e-05, "chaser_grad_norm": 0.03915706975385547, "chaser_new_fields": 94.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.894999504089355, "chaser_subspace_norm": 67.8890380859375, "collisions_per_episode": 0.675000011920929, "entropy": 0.001400032255332917, "explorer_grad_norm": 0.17385003715753555, "explorer_new_fields": 95.625, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 142.23980712890625, "final_distance": 3.8947174549102783, "max_param_abs": 2.1906604766845703, "policy_loss": -3.2025814260094876e-05, "value_loss": 0.06482306495308876}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:04.617059+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020784.6170716, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.156070570738876, "elapsed_seconds": 18561.854054659954, "env_steps_per_second": 4094.4185735002156, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2713721353020461, "eta_seconds": 976.939687087366, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00017792411029660116, "chaser_grad_norm": 0.03782106749713421, "chaser_new_fields": 95.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.970000267028809, "chaser_subspace_norm": 69.94136810302734, "collisions_per_episode": 0.4749999940395355, "entropy": 0.002133544418029487, "explorer_grad_norm": 0.33831240609288216, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 150.135009765625, "final_distance": 3.7090566158294678, "max_param_abs": 2.208369493484497, "policy_loss": -8.993363421172518e-05, "value_loss": 0.04737705644220114}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:13.929295+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021273.9293084, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.290843675516112, "elapsed_seconds": 19047.037231858005, "env_steps_per_second": 4095.1250869367486, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.135662658346567, "eta_seconds": 488.3855700476411, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0016687422878480795, "chaser_grad_norm": 0.2595461467280984, "chaser_new_fields": 94.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.909998893737793, "chaser_subspace_norm": 70.38494110107422, "collisions_per_episode": 0.9750000238418579, "entropy": 0.003381687856744975, "explorer_grad_norm": 0.2522974908351898, "explorer_new_fields": 94.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.86250305175781, "explorer_subspace_norm": 155.3250732421875, "final_distance": 3.5724456310272217, "max_param_abs": 2.1981301307678223, "policy_loss": -0.0008378539641853422, "value_loss": 0.11335896141827106}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:19.112472+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021759.1124866, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.42705643406639, "elapsed_seconds": 19537.403162639006, "env_steps_per_second": 4094.709994672293, "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-transformer/non_social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:29.478401+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022249.4784148, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.427951933531375, "elapsed_seconds": 19540.62696071295, "env_steps_per_second": 4094.034452468825, "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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:32.702201+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022252.702217, "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..c111c1b1a5aa826eb0edf920084df7dc9699870f --- /dev/null +++ b/non_social/seed_0000/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.427951933531375, + "elapsed_seconds": 19540.62696071295, + "env_steps_per_second": 4094.034452468825, + "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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0003983886290532812, + "chaser_grad_norm": 0.02647677343338728, + "chaser_new_fields": 94.9000015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.955000877380371, + "chaser_subspace_norm": 75.5232925415039, + "collisions_per_episode": 0.550000011920929, + "entropy": 0.0030606171931140125, + "explorer_grad_norm": 0.17282858304679394, + "explorer_new_fields": 95.42500305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 92.95000457763672, + "explorer_subspace_norm": 155.51156616210938, + "final_distance": 3.808248281478882, + "max_param_abs": 2.2150280475616455, + "policy_loss": -0.00044627454370527175, + "value_loss": 0.06406432762742043 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:57:32.702201+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022252.702217, + "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..c210a4bd4ee89ad23c350ee19c730c0a9b211d39 --- /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 sha256:97e55e51803fcb1f12344ef9426ee05617a44dd6420294c15ba2d71524feac2c +size 29760368 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..58bdea073646786422989721315cadbd4c43e80d --- /dev/null +++ b/non_social/seed_0001/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..0754d84f2f5467ba7ab5ae3a7c27c2b0b4ce093b --- /dev/null +++ b/non_social/seed_0001/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007073184386050949, "elapsed_seconds": 2.546346378978342, "env_steps_per_second": 1570.8781935648915, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.145661453663294, "eta_seconds": 50924.38123318786, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.006091276896187381, "chaser_grad_norm": 4.504279673099518, "chaser_new_fields": 34.10000228881836, "chaser_partner_vision": 0.0, "chaser_return": -3.2150001525878906, "chaser_subspace_norm": 0.03877149149775505, "collisions_per_episode": 1.225000023841858, "entropy": 2.748730421066284, "explorer_grad_norm": 9.367615342140198, "explorer_new_fields": 35.10000228881836, "explorer_partner_vision": 0.0, "explorer_return": 2.4625000953674316, "explorer_subspace_norm": 0.030939258635044098, "final_distance": 5.002046585083008, "max_param_abs": 1.0012006759643555, "policy_loss": -0.013167476201429551, "value_loss": 7.158067882061005}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.483055+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.483091, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13894482685805998, "elapsed_seconds": 500.2013766890159, "env_steps_per_second": 3998.389635067789, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.4188482474643385, "eta_seconds": 19507.85369087162, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0010831825758677027, "chaser_grad_norm": 0.3167974688112736, "chaser_new_fields": 90.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.170001029968262, "chaser_subspace_norm": 12.111225128173828, "collisions_per_episode": 0.574999988079071, "entropy": 0.10503689385950565, "explorer_grad_norm": 10.194164991378784, "explorer_new_fields": 88.5, "explorer_partner_vision": 0.0, "explorer_return": 82.5999984741211, "explorer_subspace_norm": 14.751652717590332, "final_distance": 3.2616829872131348, "max_param_abs": 1.3052014112472534, "policy_loss": -0.001630453222198014, "value_loss": 0.521367721259594}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:12.138077+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003212.1380937, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27813606615220327, "elapsed_seconds": 1001.2898381479317, "env_steps_per_second": 3994.8472935655973, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.284585256891862, "eta_seconds": 19024.506924810703, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.028383875501535982, "chaser_grad_norm": 1.810861423611641, "chaser_new_fields": 82.0, "chaser_partner_vision": 0.0, "chaser_return": 6.39500093460083, "chaser_subspace_norm": 17.26828384399414, "collisions_per_episode": 1.524999976158142, "entropy": 0.02800812106579542, "explorer_grad_norm": 2.9772969484329224, "explorer_new_fields": 93.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.17500305175781, "explorer_subspace_norm": 28.539579391479492, "final_distance": 2.783372640609741, "max_param_abs": 1.520729899406433, "policy_loss": 2.5617609253458795e-05, "value_loss": 0.20261814072728157}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:33.226534+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003713.2265503, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4171141794858138, "elapsed_seconds": 1501.6110461489297, "env_steps_per_second": 3995.70848615409, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.144408213658371, "eta_seconds": 18519.869569170136, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004342201401708934, "chaser_grad_norm": 0.1626635678112507, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 22.83955955505371, "collisions_per_episode": 0.5250000357627869, "entropy": 0.027943478897213936, "explorer_grad_norm": 2.787310838699341, "explorer_new_fields": 93.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.70000457763672, "explorer_subspace_norm": 35.686798095703125, "final_distance": 2.20277738571167, "max_param_abs": 1.6910988092422485, "policy_loss": -0.0016410909520914174, "value_loss": 0.17956972122192383}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:53.547740+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004213.5477538, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5539804370719422, "elapsed_seconds": 1994.329573458992, "env_steps_per_second": 4011.373098241076, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.98582393364748, "eta_seconds": 17948.966161130928, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.002541680116437073, "chaser_grad_norm": 0.11740930378437042, "chaser_new_fields": 91.875, "chaser_partner_vision": 0.0, "chaser_return": 8.315001487731934, "chaser_subspace_norm": 28.9432373046875, "collisions_per_episode": 0.800000011920929, "entropy": 0.011052386136725545, "explorer_grad_norm": 2.8440057039260864, "explorer_new_fields": 94.625, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 48.87232208251953, "final_distance": 2.915606737136841, "max_param_abs": 1.8323895931243896, "policy_loss": -0.0014348829537453867, "value_loss": 0.140011016279459}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:05:06.266271+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004706.266286, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6928622715927647, "elapsed_seconds": 2494.304177733953, "env_steps_per_second": 4009.1341261693615, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.8500359011493535, "eta_seconds": 17460.12924413767, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002890846548755377, "chaser_grad_norm": 0.054316350258886814, "chaser_new_fields": 92.5, "chaser_partner_vision": 0.0, "chaser_return": 8.444999694824219, "chaser_subspace_norm": 31.604827880859375, "collisions_per_episode": 0.824999988079071, "entropy": 0.008672359632328153, "explorer_grad_norm": 2.87642902135849, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.7874984741211, "explorer_subspace_norm": 55.69298553466797, "final_distance": 2.080470323562622, "max_param_abs": 1.9203174114227295, "policy_loss": -0.0009162586192132949, "value_loss": 0.13287755846977234}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:13:26.240872+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005206.2408843, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8267417508294521, "elapsed_seconds": 2976.2703029860277, "env_steps_per_second": 4031.891857389653, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.684869921366895, "eta_seconds": 16865.53171692082, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0026825206270671975, "chaser_grad_norm": 0.3681898396462202, "chaser_new_fields": 92.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.394999504089355, "chaser_subspace_norm": 34.616058349609375, "collisions_per_episode": 0.800000011920929, "entropy": 0.011180124478414655, "explorer_grad_norm": 1.9449609816074371, "explorer_new_fields": 94.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 90.92500305175781, "explorer_subspace_norm": 55.717529296875, "final_distance": 2.862488269805908, "max_param_abs": 1.992835521697998, "policy_loss": -0.0040723030911706815, "value_loss": 0.17838570103049278}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:21:28.206999+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005688.2070158, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9603808076341456, "elapsed_seconds": 3457.370907482924, "env_steps_per_second": 4049.31966359156, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.527509521703829, "eta_seconds": 16299.034278133784, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0023234201585415892, "chaser_grad_norm": 0.109742546454072, "chaser_new_fields": 92.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 35.43597412109375, "collisions_per_episode": 1.100000023841858, "entropy": 0.009827223373576999, "explorer_grad_norm": 0.8428630605340004, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.1875, "explorer_subspace_norm": 64.09198760986328, "final_distance": 2.3436453342437744, "max_param_abs": 2.070355176925659, "policy_loss": -0.0018424933510772235, "value_loss": 0.1317814216017723}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:29.307603+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006169.307615, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0946459529144341, "elapsed_seconds": 3940.7254304919625, "env_steps_per_second": 4060.166150170617, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.3785838116577365, "eta_seconds": 15762.90172196785, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00019805408916984124, "chaser_grad_norm": 0.05216287728399038, "chaser_new_fields": 92.25, "chaser_partner_vision": 0.0, "chaser_return": 8.355000495910645, "chaser_subspace_norm": 39.39981460571289, "collisions_per_episode": 1.149999976158142, "entropy": 0.007402043207548559, "explorer_grad_norm": 1.2529384195804596, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.7125015258789, "explorer_subspace_norm": 67.51615905761719, "final_distance": 2.6458470821380615, "max_param_abs": 2.130131244659424, "policy_loss": -0.00029325868722196446, "value_loss": 0.11367498151957989}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:32.662127+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006652.6621435, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.233532814868603, "elapsed_seconds": 4440.718133526971, "env_steps_per_second": 4053.3984501519762, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.248835251214077, "eta_seconds": 15295.806904370676, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.000477152812809134, "chaser_grad_norm": 0.09721498657017946, "chaser_new_fields": 92.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.425000190734863, "chaser_subspace_norm": 41.61264419555664, "collisions_per_episode": 1.0, "entropy": 0.006108463974669576, "explorer_grad_norm": 1.4655251801013947, "explorer_new_fields": 95.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.20000457763672, "explorer_subspace_norm": 72.17704772949219, "final_distance": 2.4420201778411865, "max_param_abs": 2.179051160812378, "policy_loss": -0.0006444988547595898, "value_loss": 0.0815439410507679}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:52.654831+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007152.6548455, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.371002308455257, "elapsed_seconds": 4935.608310438925, "env_steps_per_second": 4052.185413031974, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.113006925365771, "eta_seconds": 14806.824931316776, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.008122568582032863, "chaser_grad_norm": 0.30936599150300026, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.270000457763672, "chaser_subspace_norm": 42.52458953857422, "collisions_per_episode": 1.774999976158142, "entropy": 0.008353819604963064, "explorer_grad_norm": 1.0237028822302818, "explorer_new_fields": 93.5, "explorer_partner_vision": 0.0, "explorer_return": 89.4625015258789, "explorer_subspace_norm": 77.6466293334961, "final_distance": 2.9305949211120605, "max_param_abs": 2.235593557357788, "policy_loss": -0.001684646227046116, "value_loss": 0.23588852211833}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:54:07.545006+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007647.545027, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.5076593793275082, "elapsed_seconds": 5427.573765579029, "env_steps_per_second": 4053.3765085831087, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.974738363681612, "eta_seconds": 14309.058109253803, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0010398755102261903, "chaser_grad_norm": 0.03989175520837307, "chaser_new_fields": 92.25, "chaser_partner_vision": 0.0, "chaser_return": 8.380000114440918, "chaser_subspace_norm": 45.31040573120117, "collisions_per_episode": 1.0750000476837158, "entropy": 0.007315914612263441, "explorer_grad_norm": 0.41754402965307236, "explorer_new_fields": 94.375, "explorer_partner_vision": 0.0, "explorer_return": 91.0, "explorer_subspace_norm": 81.83265686035156, "final_distance": 2.7675256729125977, "max_param_abs": 2.2861053943634033, "policy_loss": -0.0010885612093822772, "value_loss": 0.14534751698374748}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:02:19.510460+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008139.5104733, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.647049954340554, "elapsed_seconds": 5929.379835625994, "env_steps_per_second": 4047.640843617198, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.843116560127959, "eta_seconds": 13835.219616460652, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.001582655426779822, "chaser_grad_norm": 0.08839346002787352, "chaser_new_fields": 92.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.49000072479248, "chaser_subspace_norm": 49.190311431884766, "collisions_per_episode": 0.699999988079071, "entropy": 0.003298768715467304, "explorer_grad_norm": 0.4850756637752056, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 83.96345520019531, "final_distance": 2.77482008934021, "max_param_abs": 2.3289451599121094, "policy_loss": -0.001331937531009686, "value_loss": 0.06736214831471443}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:10:41.316531+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008641.3165467, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7844666381213918, "elapsed_seconds": 6424.07989723701, "env_steps_per_second": 4047.27220332091, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7061999407136597, "eta_seconds": 13342.319786569175, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.002717901549796231, "chaser_grad_norm": 0.12501308415085077, "chaser_new_fields": 92.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.530000686645508, "chaser_subspace_norm": 48.52886199951172, "collisions_per_episode": 0.5250000357627869, "entropy": 0.0036276100436225533, "explorer_grad_norm": 0.7959117516875267, "explorer_new_fields": 95.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 85.16908264160156, "final_distance": 2.8057403564453125, "max_param_abs": 2.3636598587036133, "policy_loss": -0.0010038290538361139, "value_loss": 0.0742267556488514}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:18:56.016592+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009136.016604, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.9212862815591507, "elapsed_seconds": 6916.630613612942, "env_steps_per_second": 4048.2138723574303, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.568103094324137, "eta_seconds": 12845.171139566894, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00022324362465875325, "chaser_grad_norm": 0.024161220528185368, "chaser_new_fields": 92.875, "chaser_partner_vision": 0.0, "chaser_return": 8.52500057220459, "chaser_subspace_norm": 49.56332778930664, "collisions_per_episode": 0.5, "entropy": 0.0038169848266988993, "explorer_grad_norm": 0.2579614296555519, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 90.28097534179688, "final_distance": 2.8356008529663086, "max_param_abs": 2.406066417694092, "policy_loss": -0.000252534526921977, "value_loss": 0.05667364690452814}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:27:08.567310+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009628.567326, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.057540595391927, "elapsed_seconds": 7407.146143410937, "env_steps_per_second": 4050.1428511285208, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.429234325653212, "eta_seconds": 12345.243572351563, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.001062207278654359, "chaser_grad_norm": 0.03593328036367893, "chaser_new_fields": 92.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 51.36481475830078, "collisions_per_episode": 0.925000011920929, "entropy": 0.005192454671487212, "explorer_grad_norm": 1.0045685917139053, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 93.41552734375, "final_distance": 2.96689772605896, "max_param_abs": 2.437062978744507, "policy_loss": -0.0007657717628184457, "value_loss": 0.11875865235924721}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:35:19.082841+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010119.082855, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.195929895317208, "elapsed_seconds": 7905.347623141948, "env_steps_per_second": 4047.892834759584, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2938948429758117, "eta_seconds": 11858.021434712922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.000996393159171327, "chaser_grad_norm": 0.12287321500480175, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 51.35920333862305, "collisions_per_episode": 0.824999988079071, "entropy": 0.007126937969587743, "explorer_grad_norm": 0.5258216448128223, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 96.66453552246094, "final_distance": 2.1577203273773193, "max_param_abs": 2.472053289413452, "policy_loss": -0.0013220983569048883, "value_loss": 0.09719987586140633}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:43:37.284319+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010617.2843356, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3308561228886053, "elapsed_seconds": 8391.082042398979, "env_steps_per_second": 4051.920816433768, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1535112250845834, "eta_seconds": 11352.6404103045, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0006369755563935264, "chaser_grad_norm": 0.15209951996803284, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 55.33511734008789, "collisions_per_episode": 1.1750000715255737, "entropy": 0.003968458739109337, "explorer_grad_norm": 0.6275946795940399, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.67500305175781, "explorer_subspace_norm": 97.49080657958984, "final_distance": 2.787330150604248, "max_param_abs": 2.507138252258301, "policy_loss": -0.000677521714706586, "value_loss": 0.0855429582297802}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:51:43.018737+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011103.0187495, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.467534301508341, "elapsed_seconds": 8883.123485430027, "env_steps_per_second": 4052.6285668601463, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.0158752573990837, "eta_seconds": 10857.150926636701, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 6.216459688701548e-05, "chaser_grad_norm": 0.0843466967344284, "chaser_new_fields": 94.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.745000839233398, "chaser_subspace_norm": 53.2131233215332, "collisions_per_episode": 0.7250000238418579, "entropy": 0.007430837024003267, "explorer_grad_norm": 0.35458748042583466, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 98.6747817993164, "final_distance": 2.9391210079193115, "max_param_abs": 2.5464768409729004, "policy_loss": -0.0005818257509364599, "value_loss": 0.08211887069046497}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:59:55.060182+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011595.0601966, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.604981942926954, "elapsed_seconds": 9377.934994537034, "env_steps_per_second": 4052.0647692841007, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.879190568498212, "eta_seconds": 10365.086046593564, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0010052093184430566, "chaser_grad_norm": 0.09301582723855972, "chaser_new_fields": 94.125, "chaser_partner_vision": 0.0, "chaser_return": 8.75, "chaser_subspace_norm": 56.764808654785156, "collisions_per_episode": 0.949999988079071, "entropy": 0.0033797190408222377, "explorer_grad_norm": 0.5405961982905865, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.5625, "explorer_subspace_norm": 102.1544189453125, "final_distance": 2.845025062561035, "max_param_abs": 2.579113483428955, "policy_loss": -0.0005198264389183471, "value_loss": 0.0823406670242548}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:08:09.871691+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012089.8717058, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7412828842372012, "elapsed_seconds": 9868.618383253925, "env_steps_per_second": 4053.2522838127034, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7412828842372012, "eta_seconds": 9868.618383253925, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0023042803507682397, "chaser_grad_norm": 0.04860458383336663, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 59.11008834838867, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0030235633021220565, "explorer_grad_norm": 0.4425838142633438, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 102.14756774902344, "final_distance": 2.8141345977783203, "max_param_abs": 2.6225476264953613, "policy_loss": -0.0009095669351530944, "value_loss": 0.09786444157361984}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:16:20.555079+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012580.555093, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.879656619975544, "elapsed_seconds": 10366.763831911958, "env_steps_per_second": 4051.4089720758957, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.605403608549302, "eta_seconds": 9379.452990777487, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005118195990925223, "chaser_grad_norm": 0.08334831520915031, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 59.05033874511719, "collisions_per_episode": 0.800000011920929, "entropy": 0.003657435707282275, "explorer_grad_norm": 0.2816202864050865, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.3499984741211, "explorer_subspace_norm": 106.18612670898438, "final_distance": 2.3185527324676514, "max_param_abs": 2.6511926651000977, "policy_loss": -0.0004923692047587025, "value_loss": 0.07210455648601055}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:24:38.700526+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013078.7005394, "update": 10500} +{"device": "cuda", "elapsed_hours": 3.0160725797041477, "elapsed_seconds": 10857.86128693493, "env_steps_per_second": 4052.363429338005, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.467695747030666, "eta_seconds": 8883.704689310398, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00017792440830510792, "chaser_grad_norm": 0.04437321983277798, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.795000076293945, "chaser_subspace_norm": 57.75242233276367, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0032481419038958848, "explorer_grad_norm": 0.24263937771320343, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 102.43596649169922, "final_distance": 2.4457814693450928, "max_param_abs": 2.6756579875946045, "policy_loss": -0.00031847636937154533, "value_loss": 0.06373031809926033}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:32:49.797984+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013569.797999, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.155583950798868, "elapsed_seconds": 11360.102222875925, "env_steps_per_second": 4049.259337417708, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3323881375469893, "eta_seconds": 8396.597295169162, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.008585023168623529, "chaser_grad_norm": 0.234678759239614, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 58.88050079345703, "collisions_per_episode": 0.6000000238418579, "entropy": 0.005969225661829114, "explorer_grad_norm": 0.48086490482091904, "explorer_new_fields": 94.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.63750457763672, "explorer_subspace_norm": 100.81809997558594, "final_distance": 2.4236786365509033, "max_param_abs": 2.7097814083099365, "policy_loss": -0.004882263489065508, "value_loss": 0.20253624022006989}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:41:12.038918+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014072.0389283, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.289983558591662, "elapsed_seconds": 11843.940810929984, "env_steps_per_second": 4052.7051566910905, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1933223723944413, "eta_seconds": 7895.960540619989, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00019440097922895916, "chaser_grad_norm": 0.029068120289593935, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.894999504089355, "chaser_subspace_norm": 62.05281448364258, "collisions_per_episode": 0.625, "entropy": 0.003293231420684606, "explorer_grad_norm": 0.493771456182003, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 110.28379821777344, "final_distance": 2.185750961303711, "max_param_abs": 2.739060878753662, "policy_loss": -0.0002078625110115695, "value_loss": 0.06744680739939213}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:49:15.877507+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014555.8775215, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.4264354140583437, "elapsed_seconds": 12335.167490610038, "env_steps_per_second": 4053.4512432086353, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0558612484350065, "eta_seconds": 7401.100494366023, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005562198738733315, "chaser_grad_norm": 0.08132471237331629, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.86500072479248, "chaser_subspace_norm": 60.91058349609375, "collisions_per_episode": 0.625, "entropy": 0.0040358962723985314, "explorer_grad_norm": 0.3306007981300354, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 111.05867767333984, "final_distance": 2.458824634552002, "max_param_abs": 2.7716782093048096, "policy_loss": -0.0005716206628125775, "value_loss": 0.06918930821120739}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:57:27.104189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015047.104203, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.563489204867219, "elapsed_seconds": 12828.561137521989, "env_steps_per_second": 4053.4553674851572, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9188018795438873, "eta_seconds": 6907.686766357994, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00011050143215055735, "chaser_grad_norm": 0.049098674207925797, "chaser_new_fields": 94.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.920000076293945, "chaser_subspace_norm": 63.42983627319336, "collisions_per_episode": 0.699999988079071, "entropy": 0.003923855605535209, "explorer_grad_norm": 0.3020031377673149, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 110.25255584716797, "final_distance": 2.292327404022217, "max_param_abs": 2.8057360649108887, "policy_loss": -0.0004508834210561474, "value_loss": 0.04780322127044201}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:05:40.497834+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015540.4978487, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.69906137363028, "elapsed_seconds": 13316.620945069008, "env_steps_per_second": 4055.0827588131942, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.781029550266431, "eta_seconds": 6411.706380959152, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00023009996807599187, "chaser_grad_norm": 0.06983139086514711, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.8100004196167, "chaser_subspace_norm": 63.69921112060547, "collisions_per_episode": 0.699999988079071, "entropy": 0.003970191464759409, "explorer_grad_norm": 0.6755344867706299, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 115.0482406616211, "final_distance": 2.1381518840789795, "max_param_abs": 2.828923225402832, "policy_loss": -0.0004508252339432417, "value_loss": 0.06965262070298195}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:13:48.557642+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016028.557656, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.835262262045257, "elapsed_seconds": 13806.944143362925, "env_steps_per_second": 4055.930075368597, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6436838265908245, "eta_seconds": 5917.261775726968, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 7.470461521874583e-05, "chaser_grad_norm": 0.07316626608371735, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 67.3201675415039, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0024352622567676008, "explorer_grad_norm": 0.40397659689188004, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 119.69538116455078, "final_distance": 2.3341729640960693, "max_param_abs": 2.8383781909942627, "policy_loss": -7.460785411250725e-05, "value_loss": 0.09181175008416176}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:21:58.880839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016518.880855, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.970110981043318, "elapsed_seconds": 14292.399531755946, "env_steps_per_second": 4058.100941771965, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5059041652233276, "eta_seconds": 5421.25499480398, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 2.7242026847673628e-05, "chaser_grad_norm": 0.0497783487662673, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.850000381469727, "chaser_subspace_norm": 67.51028442382812, "collisions_per_episode": 0.5250000357627869, "entropy": 0.0024516364210285246, "explorer_grad_norm": 0.25298214703798294, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 121.70927429199219, "final_distance": 2.5439205169677734, "max_param_abs": 2.869907855987549, "policy_loss": -6.890869436126934e-05, "value_loss": 0.05644155852496624}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:30:04.336228+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017004.3362422, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.107270685498596, "elapsed_seconds": 14786.174467794946, "env_steps_per_second": 4057.844720463911, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3690902284995319, "eta_seconds": 4928.724822598315, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0006331868479183891, "chaser_grad_norm": 0.04258312797173858, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.51932525634766, "collisions_per_episode": 0.375, "entropy": 0.0030604162020608783, "explorer_grad_norm": 0.4632948637008667, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 122.46908569335938, "final_distance": 2.23240065574646, "max_param_abs": 2.899177074432373, "policy_loss": -0.00021183396170698643, "value_loss": 0.04441228974610567}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:38:18.111163+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017498.1111782, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.24289645000694, "elapsed_seconds": 15274.427220024983, "env_steps_per_second": 4059.072010158073, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2318086467762082, "eta_seconds": 4434.51112839435, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0013615168079588735, "chaser_grad_norm": 0.04086262080818415, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.875, "chaser_subspace_norm": 69.30021667480469, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0028082082280889153, "explorer_grad_norm": 0.1993851512670517, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.7249984741211, "explorer_subspace_norm": 119.8309555053711, "final_distance": 2.0393521785736084, "max_param_abs": 2.918504238128662, "policy_loss": -0.00019859792260112918, "value_loss": 0.04485477786511183}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:46:26.363915+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017986.3639271, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.377337626529431, "elapsed_seconds": 15758.41545550595, "env_steps_per_second": 4061.322039687598, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0943344066323577, "eta_seconds": 3939.6038638764876, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006552253194327395, "chaser_grad_norm": 0.026540640741586685, "chaser_new_fields": 94.75, "chaser_partner_vision": 0.0, "chaser_return": 8.905000686645508, "chaser_subspace_norm": 69.47985076904297, "collisions_per_episode": 0.45000001788139343, "entropy": 0.004001817433163524, "explorer_grad_norm": 0.2520532086491585, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.5374984741211, "explorer_subspace_norm": 132.6637725830078, "final_distance": 2.078486680984497, "max_param_abs": 2.937121629714966, "policy_loss": -0.0010853266977051135, "value_loss": 0.08242643252015114}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:54:30.352153+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018470.3521686, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.5111880817922065, "elapsed_seconds": 16240.277094451943, "env_steps_per_second": 4063.970067514866, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9569186840165287, "eta_seconds": 3444.9072624595033, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0011654459594116588, "chaser_grad_norm": 0.12811908777803183, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 72.75859069824219, "collisions_per_episode": 1.475000023841858, "entropy": 0.0036797402426600456, "explorer_grad_norm": 0.2041437365114689, "explorer_new_fields": 94.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 129.2154998779297, "final_distance": 2.415665864944458, "max_param_abs": 2.957319498062134, "policy_loss": -0.0006939259038780676, "value_loss": 0.2137848101556301}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:02:32.213790+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018952.2138023, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.6476044810977575, "elapsed_seconds": 16731.376131951925, "env_steps_per_second": 4064.220388312252, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.82016549666431, "eta_seconds": 2952.5957879915163, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 5.5648104003269694e-05, "chaser_grad_norm": 0.018644486088305712, "chaser_new_fields": 94.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.880000114440918, "chaser_subspace_norm": 72.43238067626953, "collisions_per_episode": 0.7250000238418579, "entropy": 0.001984165806788951, "explorer_grad_norm": 0.2640278488397598, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 126.32189178466797, "final_distance": 2.4246740341186523, "max_param_abs": 2.9786489009857178, "policy_loss": -5.09057059243645e-05, "value_loss": 0.043034049682319164}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:10:43.312827+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019443.3128421, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.785597953629169, "elapsed_seconds": 17228.15263306501, "env_steps_per_second": 4063.117009170966, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6836568505184527, "eta_seconds": 2461.16466186643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 2.6341494720749647e-05, "chaser_grad_norm": 0.07059159502387047, "chaser_new_fields": 94.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.920001029968262, "chaser_subspace_norm": 75.26416778564453, "collisions_per_episode": 0.42500001192092896, "entropy": 0.002113209222443402, "explorer_grad_norm": 0.316756010055542, "explorer_new_fields": 95.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 127.33657836914062, "final_distance": 2.3594603538513184, "max_param_abs": 2.989992141723633, "policy_loss": -4.8084022527206116e-05, "value_loss": 0.026043227408081293}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:19:00.089328+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019940.089339, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.921170455568612, "elapsed_seconds": 17716.213640047004, "env_steps_per_second": 4064.0738175140323, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5467967172854014, "eta_seconds": 1968.468182227445, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0008898909904117291, "chaser_grad_norm": 0.042340966407209635, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 72.6676254272461, "collisions_per_episode": 0.800000011920929, "entropy": 0.001199792983243242, "explorer_grad_norm": 0.21852006390690804, "explorer_new_fields": 95.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 124.67669677734375, "final_distance": 2.4179787635803223, "max_param_abs": 3.023815155029297, "policy_loss": -0.0005869875105570799, "value_loss": 0.08416097238659859}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:27:08.150335+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020428.1503508, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.057042567851927, "elapsed_seconds": 18205.35324426694, "env_steps_per_second": 4064.7384869229822, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4100304784744806, "eta_seconds": 1476.10972250813, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 9.724407010380798e-05, "chaser_grad_norm": 0.053080628626048565, "chaser_new_fields": 94.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.905000686645508, "chaser_subspace_norm": 78.34722137451172, "collisions_per_episode": 0.5250000357627869, "entropy": 0.0021247865515761077, "explorer_grad_norm": 0.20850447192788124, "explorer_new_fields": 95.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 131.81642150878906, "final_distance": 2.267756223678589, "max_param_abs": 3.0450665950775146, "policy_loss": -0.0002477926192216984, "value_loss": 0.03876691497862339}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:35:17.289939+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020917.289952, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.1930706142980485, "elapsed_seconds": 18695.054211472976, "env_steps_per_second": 4065.2463020599066, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2733195060156868, "eta_seconds": 983.9502216564724, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0028486762956183576, "chaser_grad_norm": 0.13814295083284378, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.760001182556152, "chaser_subspace_norm": 75.9752197265625, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0027867945027537644, "explorer_grad_norm": 0.16607102751731873, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 136.76522827148438, "final_distance": 2.363304853439331, "max_param_abs": 3.070605754852295, "policy_loss": -0.0013202877221658582, "value_loss": 0.09105210937559605}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:43:26.990907+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021406.990922, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.327256802463316, "elapsed_seconds": 19178.124488867936, "env_steps_per_second": 4067.1338871157914, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13659632826829013, "eta_seconds": 491.7467817658445, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0014851364632712527, "chaser_grad_norm": 0.08397708088159561, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 80.55755615234375, "collisions_per_episode": 0.800000011920929, "entropy": 0.0034964141086675227, "explorer_grad_norm": 0.2840192373842001, "explorer_new_fields": 95.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 135.56906127929688, "final_distance": 2.4301421642303467, "max_param_abs": 3.0977866649627686, "policy_loss": -0.0005637188296532258, "value_loss": 0.09781094640493393}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:51:30.061184+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021890.0611954, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.456585522624987, "elapsed_seconds": 19643.707881449955, "env_steps_per_second": 4072.550889210993, "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-transformer/non_social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:59:15.644576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022355.6445875, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.457300321458897, "elapsed_seconds": 19646.28115725203, "env_steps_per_second": 4072.0174652733, "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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:59:18.217854+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022358.2178695, "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..a6911b7f250343087868f6d1dc54568dfb2bee40 --- /dev/null +++ b/non_social/seed_0001/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.457300321458897, + "elapsed_seconds": 19646.28115725203, + "env_steps_per_second": 4072.0174652733, + "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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0031268233263772593, + "chaser_grad_norm": 0.2260504849255085, + "chaser_new_fields": 93.4000015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.640000343322754, + "chaser_subspace_norm": 81.0290298461914, + "collisions_per_episode": 1.149999976158142, + "entropy": 0.003422852954827249, + "explorer_grad_norm": 0.24985111132264137, + "explorer_new_fields": 94.82500457763672, + "explorer_partner_vision": 0.0, + "explorer_return": 91.82500457763672, + "explorer_subspace_norm": 135.97732543945312, + "final_distance": 2.2879505157470703, + "max_param_abs": 3.1244027614593506, + "policy_loss": -0.0035095140663869984, + "value_loss": 0.2153652273118496 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:59:18.217854+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022358.2178695, + "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..09f363ddae828e6bedb83b95d5f9df21c63fc9c5 --- /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 sha256:6b05e3f4ca6b8f514ca68a393b1bdc046876c2fdb9517fb51b9942361d3a57ac +size 29760368 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..a4af50c105ceab5125041cbd0d5c467ab691b3d7 --- /dev/null +++ b/non_social/seed_0002/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..ca4757518a5f358c09f392d3d8ebdf664cca36c9 --- /dev/null +++ b/non_social/seed_0002/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007321728369182286, "elapsed_seconds": 2.635822212905623, "env_steps_per_second": 1517.5530354115056, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.642724565527654, "eta_seconds": 52713.808435899555, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0061279205385336155, "chaser_grad_norm": 8.69063800573349, "chaser_new_fields": 32.35000228881836, "chaser_partner_vision": 0.0, "chaser_return": -3.5600006580352783, "chaser_subspace_norm": 0.034794267266988754, "collisions_per_episode": 1.6750000715255737, "entropy": 2.7557111382484436, "explorer_grad_norm": 6.052844345569611, "explorer_new_fields": 35.82500076293945, "explorer_partner_vision": 0.0, "explorer_return": 3.4375, "explorer_subspace_norm": 0.031101124361157417, "final_distance": 5.287587642669678, "max_param_abs": 1.0012022256851196, "policy_loss": -0.01202418064700339, "value_loss": 7.6772836446762085}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.814566+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.8145802, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13635041027666173, "elapsed_seconds": 490.86147699598223, "env_steps_per_second": 4074.4692621628774, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.317666000789807, "eta_seconds": 19143.597602843307, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0014374513662449573, "chaser_grad_norm": 0.22422203421592712, "chaser_new_fields": 90.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.045001029968262, "chaser_subspace_norm": 11.34537124633789, "collisions_per_episode": 0.949999988079071, "entropy": 0.07179526425898075, "explorer_grad_norm": 10.016987323760986, "explorer_new_fields": 92.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 88.1875, "explorer_subspace_norm": 16.648193359375, "final_distance": 2.3869614601135254, "max_param_abs": 1.259480357170105, "policy_loss": -0.0028794059618419965, "value_loss": 0.48020658642053604}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:03.040222+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003203.0402334, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27326310440720086, "elapsed_seconds": 983.747175865923, "env_steps_per_second": 4066.085370439903, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.191998983736816, "eta_seconds": 18691.196341452538, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0014687455615272071, "chaser_grad_norm": 0.20035238936543465, "chaser_new_fields": 93.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.530000686645508, "chaser_subspace_norm": 14.68276309967041, "collisions_per_episode": 1.75, "entropy": 0.028629083652049303, "explorer_grad_norm": 5.136371970176697, "explorer_new_fields": 94.5, "explorer_partner_vision": 0.0, "explorer_return": 90.8125, "explorer_subspace_norm": 29.166044235229492, "final_distance": 2.825841188430786, "max_param_abs": 1.3679320812225342, "policy_loss": -0.0019738255795118675, "value_loss": 0.426460824906826}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:15.925921+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003695.9259365, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.40687836955388773, "elapsed_seconds": 1464.762130393996, "env_steps_per_second": 4096.228237677133, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.018166557831282, "eta_seconds": 18065.399608192616, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0012405988440074012, "chaser_grad_norm": 0.11184354312717915, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.59000015258789, "chaser_subspace_norm": 16.197612762451172, "collisions_per_episode": 1.5750000476837158, "entropy": 0.020140967797487974, "explorer_grad_norm": 4.171184957027435, "explorer_new_fields": 96.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.13750457763672, "explorer_subspace_norm": 39.99115753173828, "final_distance": 2.9582443237304688, "max_param_abs": 1.439399003982544, "policy_loss": -0.0013640758586177526, "value_loss": 0.2158144824206829}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:16.940873+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004176.9408846, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5437295976183183, "elapsed_seconds": 1957.426551425946, "env_steps_per_second": 4086.9988169784256, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.893566378564865, "eta_seconds": 17616.838962833514, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0013608692952107315, "chaser_grad_norm": 0.10284747928380966, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 18.577129364013672, "collisions_per_episode": 1.850000023841858, "entropy": 0.009836691664531827, "explorer_grad_norm": 2.7826343774795532, "explorer_new_fields": 96.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.4375, "explorer_subspace_norm": 51.28536605834961, "final_distance": 2.7944366931915283, "max_param_abs": 1.4927884340286255, "policy_loss": -0.0007597444546405541, "value_loss": 0.19846593216061592}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:29.605296+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004669.6053092, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6781449054674724, "elapsed_seconds": 2441.3216596829006, "env_steps_per_second": 4096.141923919556, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.747014338272307, "eta_seconds": 17089.251617780305, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00016304003913525378, "chaser_grad_norm": 0.0689720269292593, "chaser_new_fields": 93.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.545000076293945, "chaser_subspace_norm": 21.22562026977539, "collisions_per_episode": 2.049999952316284, "entropy": 0.011092229979112744, "explorer_grad_norm": 3.2772421091794968, "explorer_new_fields": 96.625, "explorer_partner_vision": 0.0, "explorer_return": 93.70000457763672, "explorer_subspace_norm": 56.04767608642578, "final_distance": 2.6179006099700928, "max_param_abs": 1.5319234132766724, "policy_loss": -0.00040609817267522885, "value_loss": 0.287984736263752}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:33.500402+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005153.5004132, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8128366362458069, "elapsed_seconds": 2926.211890484905, "env_steps_per_second": 4100.865025878721, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.606074272059572, "eta_seconds": 16581.86737941446, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0006030650421495198, "chaser_grad_norm": 0.07318039052188396, "chaser_new_fields": 93.75, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 24.370473861694336, "collisions_per_episode": 1.5500000715255737, "entropy": 0.006501625059172511, "explorer_grad_norm": 1.3841366171836853, "explorer_new_fields": 96.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.07500457763672, "explorer_subspace_norm": 62.4726676940918, "final_distance": 3.1512153148651123, "max_param_abs": 1.56986665725708, "policy_loss": -0.0006041124149902544, "value_loss": 0.20404154062271118}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:38.390636+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005638.3906512, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9494057591094235, "elapsed_seconds": 3417.8607327939244, "env_steps_per_second": 4096.129448947946, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.475770007230139, "eta_seconds": 16112.7720260285, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 5.0752096209590036e-05, "chaser_grad_norm": 0.023512616753578186, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.580000877380371, "chaser_subspace_norm": 25.51074981689453, "collisions_per_episode": 1.5500000715255737, "entropy": 0.007038616458885372, "explorer_grad_norm": 1.4274124652147293, "explorer_new_fields": 96.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.3375015258789, "explorer_subspace_norm": 66.74330139160156, "final_distance": 2.7487001419067383, "max_param_abs": 1.606393575668335, "policy_loss": -0.0008341384197114721, "value_loss": 0.16259713098406792}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:50.039478+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006130.039492, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0876443207569213, "elapsed_seconds": 3915.519554724917, "env_steps_per_second": 4086.3031780016418, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.350577283027685, "eta_seconds": 15662.078218899667, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0017131934787162212, "chaser_grad_norm": 0.0729985972866416, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 26.832889556884766, "collisions_per_episode": 1.75, "entropy": 0.008575530722737312, "explorer_grad_norm": 1.4874772131443024, "explorer_new_fields": 96.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.8125, "explorer_subspace_norm": 71.17113494873047, "final_distance": 3.008681535720825, "max_param_abs": 1.636663556098938, "policy_loss": -0.0019146683402411213, "value_loss": 0.19464126974344254}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:07.698298+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006627.698313, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2238280326572002, "elapsed_seconds": 4405.78091756592, "env_steps_per_second": 4085.541323272273, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.2154076680414665, "eta_seconds": 15175.467604949281, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0012005009262245625, "chaser_grad_norm": 0.11612079106271267, "chaser_new_fields": 93.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.645001411437988, "chaser_subspace_norm": 27.53525733947754, "collisions_per_episode": 1.600000023841858, "entropy": 0.004619026556611061, "explorer_grad_norm": 1.6260186731815338, "explorer_new_fields": 96.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.07500457763672, "explorer_subspace_norm": 77.10869598388672, "final_distance": 2.85575795173645, "max_param_abs": 1.6697918176651, "policy_loss": -0.0011311787646235416, "value_loss": 0.18725960701704025}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:17.959662+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007117.9596765, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3576553877099973, "elapsed_seconds": 4887.55939575599, "env_steps_per_second": 4092.0218826121236, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.072966163129992, "eta_seconds": 14662.678187267971, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005470263573985612, "chaser_grad_norm": 0.08314792811870575, "chaser_new_fields": 93.0, "chaser_partner_vision": 0.0, "chaser_return": 8.460000038146973, "chaser_subspace_norm": 30.84358787536621, "collisions_per_episode": 1.5500000715255737, "entropy": 0.007231791387312114, "explorer_grad_norm": 2.956935614347458, "explorer_new_fields": 96.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.3375015258789, "explorer_subspace_norm": 82.12212371826172, "final_distance": 2.6573290824890137, "max_param_abs": 1.700556755065918, "policy_loss": -0.0010330596112169133, "value_loss": 0.18062416464090347}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:19.738140+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007599.7381537, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4912110790266564, "elapsed_seconds": 5368.359884495963, "env_steps_per_second": 4098.085909541362, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.931374662888458, "eta_seconds": 14152.948786398449, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.001206373677458461, "chaser_grad_norm": 0.08986116107553244, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 32.25676727294922, "collisions_per_episode": 1.4250000715255737, "entropy": 0.0041676408145576715, "explorer_grad_norm": 0.5372973829507828, "explorer_new_fields": 96.75, "explorer_partner_vision": 0.0, "explorer_return": 94.3375015258789, "explorer_subspace_norm": 85.73692321777344, "final_distance": 2.9968020915985107, "max_param_abs": 1.729843020439148, "policy_loss": -0.0007324595502566922, "value_loss": 0.14542652666568756}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:20.538629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008080.5386395, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.625052962143317, "elapsed_seconds": 5850.190663715941, "env_steps_per_second": 4102.430395790829, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.791790245001073, "eta_seconds": 13650.444882003863, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0006094975457799912, "chaser_grad_norm": 0.05917355325073004, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 33.52567672729492, "collisions_per_episode": 1.5, "entropy": 0.00545320357196033, "explorer_grad_norm": 0.6449002400040627, "explorer_new_fields": 96.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.5999984741211, "explorer_subspace_norm": 85.36280822753906, "final_distance": 2.549459934234619, "max_param_abs": 1.7547281980514526, "policy_loss": -0.0006153682163017038, "value_loss": 0.12353493086993694}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:22.369408+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008562.3694234, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7589124985274975, "elapsed_seconds": 6332.084994698991, "env_steps_per_second": 4106.072489830179, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.653125958480187, "eta_seconds": 13151.253450528673, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00042492411724981416, "chaser_grad_norm": 0.04196750558912754, "chaser_new_fields": 94.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.745000839233398, "chaser_subspace_norm": 35.55250930786133, "collisions_per_episode": 1.0500000715255737, "entropy": 0.004787102923728526, "explorer_grad_norm": 1.0047070682048798, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 89.75172424316406, "final_distance": 3.230062246322632, "max_param_abs": 1.7901263236999512, "policy_loss": -0.000798380141378896, "value_loss": 0.10945381037890911}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:24.263738+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009044.263749, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8928185786483098, "elapsed_seconds": 6814.146883133915, "env_steps_per_second": 4109.098391950488, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5152345032040038, "eta_seconds": 12654.844211534413, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00019714361857248175, "chaser_grad_norm": 0.03700845502316952, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 36.750267028808594, "collisions_per_episode": 1.600000023841858, "entropy": 0.0030598515877500176, "explorer_grad_norm": 1.1551428139209747, "explorer_new_fields": 97.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.63750457763672, "explorer_subspace_norm": 88.27851104736328, "final_distance": 3.290786027908325, "max_param_abs": 1.8184542655944824, "policy_loss": -0.0002503604961097139, "value_loss": 0.1453133225440979}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:26.325628+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009526.3256426, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0287750570827563, "elapsed_seconds": 7303.590205497923, "env_steps_per_second": 4107.568901855542, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.381291761804594, "eta_seconds": 12172.65034249654, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0006325586149842077, "chaser_grad_norm": 0.07497241348028183, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 38.45086669921875, "collisions_per_episode": 0.9750000238418579, "entropy": 0.004371156450361013, "explorer_grad_norm": 1.1777122020721436, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 98.05054473876953, "final_distance": 2.680711030960083, "max_param_abs": 1.8433010578155518, "policy_loss": -0.0005633578459964816, "value_loss": 0.09415272064507008}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:35.768948+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010015.7689602, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.165660841056637, "elapsed_seconds": 7796.379027803894, "env_steps_per_second": 4104.469508970737, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.248491261584956, "eta_seconds": 11694.568541705841, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00016110613666535205, "chaser_grad_norm": 0.05154063366353512, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 38.86147689819336, "collisions_per_episode": 0.925000011920929, "entropy": 0.004242159542627633, "explorer_grad_norm": 1.4051325917243958, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 96.03810119628906, "final_distance": 3.4580161571502686, "max_param_abs": 1.8778479099273682, "policy_loss": -0.000521032350832229, "value_loss": 0.09992639347910881}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:41:48.557774+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010508.5577881, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3019769951799796, "elapsed_seconds": 8287.117182647926, "env_steps_per_second": 4102.753617529542, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1144394640670305, "eta_seconds": 11211.98207064131, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.00018751345209144787, "chaser_grad_norm": 0.04086209274828434, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.74000072479248, "chaser_subspace_norm": 41.277713775634766, "collisions_per_episode": 1.350000023841858, "entropy": 0.00669125746935606, "explorer_grad_norm": 0.7252416163682938, "explorer_new_fields": 96.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.5250015258789, "explorer_subspace_norm": 99.34879302978516, "final_distance": 2.824924945831299, "max_param_abs": 1.9064655303955078, "policy_loss": -0.0007197263473628368, "value_loss": 0.11501258239150047}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:49:59.295926+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010999.295938, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.439056712242212, "elapsed_seconds": 8780.604164071963, "env_steps_per_second": 4099.9456674408575, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9810693149627037, "eta_seconds": 10731.849533865734, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0008318895781759261, "chaser_grad_norm": 0.029094298603013158, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 41.468994140625, "collisions_per_episode": 1.274999976158142, "entropy": 0.003884760953951627, "explorer_grad_norm": 0.7414844408631325, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 103.50614929199219, "final_distance": 3.1373450756073, "max_param_abs": 1.942590355873108, "policy_loss": -0.0003571343062773735, "value_loss": 0.12418003007769585}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:12.782910+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011492.7829242, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.573374887945829, "elapsed_seconds": 9264.149596604984, "env_steps_per_second": 4101.833590200852, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8442564550980216, "eta_seconds": 10239.323238352878, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 2.222422159698506e-05, "chaser_grad_norm": 0.02208172483369708, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.795000076293945, "chaser_subspace_norm": 44.127201080322266, "collisions_per_episode": 1.25, "entropy": 0.0022298935218714178, "explorer_grad_norm": 0.3783605508506298, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 94.7125015258789, "explorer_subspace_norm": 100.45281219482422, "final_distance": 3.419586181640625, "max_param_abs": 1.9667505025863647, "policy_loss": -4.9121858756162595e-05, "value_loss": 0.10526370443403721}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:16.328340+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011976.3283513, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7105219278494284, "elapsed_seconds": 9757.878940257942, "env_steps_per_second": 4099.25151202405, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7105219278494284, "eta_seconds": 9757.878940257942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 9.216201643482552e-06, "chaser_grad_norm": 0.03743532067164779, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.770000457763672, "chaser_subspace_norm": 45.39440155029297, "collisions_per_episode": 1.225000023841858, "entropy": 0.002705720718950033, "explorer_grad_norm": 0.6737443953752518, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 103.60910034179688, "final_distance": 3.6136252880096436, "max_param_abs": 1.9912539720535278, "policy_loss": -0.00025387931091458427, "value_loss": 0.11213633045554161}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:30.057687+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012470.0577025, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8467568718997063, "elapsed_seconds": 10248.324738838943, "env_steps_per_second": 4098.230790914446, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.57563716981402, "eta_seconds": 9272.293811330472, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00021304979432612559, "chaser_grad_norm": 0.04589147632941604, "chaser_new_fields": 94.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.75, "chaser_subspace_norm": 42.61776351928711, "collisions_per_episode": 1.5, "entropy": 0.004361974424682558, "explorer_grad_norm": 0.6482382416725159, "explorer_new_fields": 96.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.48750305175781, "explorer_subspace_norm": 104.19003295898438, "final_distance": 3.3916919231414795, "max_param_abs": 2.0091161727905273, "policy_loss": -0.0011383400935436327, "value_loss": 0.17632096633315086}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:22:40.503482+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012960.5034928, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9841719255210934, "elapsed_seconds": 10743.018931875937, "env_steps_per_second": 4095.6829992588273, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4415952117899855, "eta_seconds": 8789.742762443948, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 1.509851393197148e-05, "chaser_grad_norm": 0.03615148039534688, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.780000686645508, "chaser_subspace_norm": 43.497840881347656, "collisions_per_episode": 1.149999976158142, "entropy": 0.0033954427344724536, "explorer_grad_norm": 0.25947869941592216, "explorer_new_fields": 97.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 103.66839599609375, "final_distance": 3.3145911693573, "max_param_abs": 2.024752378463745, "policy_loss": -3.67707628541325e-05, "value_loss": 0.13062887638807297}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:30:55.197678+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013455.1976929, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1180544764244242, "elapsed_seconds": 11224.996115127928, "env_steps_per_second": 4097.996963937101, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3046489608354443, "eta_seconds": 8296.736259007599, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0022483308288680365, "chaser_grad_norm": 0.039612220134586096, "chaser_new_fields": 94.25, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 45.17502975463867, "collisions_per_episode": 1.125, "entropy": 0.0031940711778588593, "explorer_grad_norm": 0.32754963636398315, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 108.3421859741211, "final_distance": 3.448914051055908, "max_param_abs": 2.0493438243865967, "policy_loss": -0.001800878806509587, "value_loss": 0.1461172215640545}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:38:57.174859+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013937.1748703, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2521516086605455, "elapsed_seconds": 11707.745791177964, "env_steps_per_second": 4099.849864879115, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1681010724403635, "eta_seconds": 7805.163860785309, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.000790681346329869, "chaser_grad_norm": 0.04079975187778473, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 46.499290466308594, "collisions_per_episode": 1.0750000476837158, "entropy": 0.004843202885240316, "explorer_grad_norm": 0.3729347884654999, "explorer_new_fields": 97.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 111.02171325683594, "final_distance": 3.2637646198272705, "max_param_abs": 2.065484046936035, "policy_loss": -0.0003173275126648889, "value_loss": 0.08790707774460316}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:46:59.924537+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014419.9245532, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3880419566197086, "elapsed_seconds": 12196.951043830952, "env_steps_per_second": 4099.385151282484, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0328251739718253, "eta_seconds": 7318.170626298572, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.003345426901507409, "chaser_grad_norm": 0.08309758501127362, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.760001182556152, "chaser_subspace_norm": 46.250267028808594, "collisions_per_episode": 1.0500000715255737, "entropy": 0.003117095388006419, "explorer_grad_norm": 0.24165698513388634, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 111.1562271118164, "final_distance": 3.4718072414398193, "max_param_abs": 2.08314847946167, "policy_loss": -0.00027372838303563185, "value_loss": 0.08545360900461674}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:09.129788+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014909.1297987, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5235442068249943, "elapsed_seconds": 12684.75914456998, "env_steps_per_second": 4099.407754404218, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8972930344442278, "eta_seconds": 6830.25492399922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 6.795539224679813e-05, "chaser_grad_norm": 0.022474604193121195, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 46.87422561645508, "collisions_per_episode": 0.949999988079071, "entropy": 0.0026716027641668916, "explorer_grad_norm": 0.40012921392917633, "explorer_new_fields": 97.375, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 113.71060180664062, "final_distance": 3.3464126586914062, "max_param_abs": 2.1162638664245605, "policy_loss": -0.00010689926428320007, "value_loss": 0.0879583116620779}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:16.937891+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015396.9379056, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.659062793856946, "elapsed_seconds": 13172.626057885005, "env_steps_per_second": 4099.410380489479, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.761770974820011, "eta_seconds": 6342.375509352039, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0037481113682407818, "chaser_grad_norm": 0.04439642280340195, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.755000114440918, "chaser_subspace_norm": 46.802284240722656, "collisions_per_episode": 1.4500000476837158, "entropy": 0.0057298337342217565, "explorer_grad_norm": 0.5554292425513268, "explorer_new_fields": 96.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.73750305175781, "explorer_subspace_norm": 119.93209075927734, "final_distance": 3.259939670562744, "max_param_abs": 2.131298303604126, "policy_loss": -0.0031371804652593482, "value_loss": 0.17059676349163055}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:24.804803+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015884.8048182, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7951653686480453, "elapsed_seconds": 13662.595327132964, "env_steps_per_second": 4098.782014628501, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6264994437063054, "eta_seconds": 5855.3979973426995, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0010495001250528935, "chaser_grad_norm": 0.025360917206853628, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 49.11723709106445, "collisions_per_episode": 1.2000000476837158, "entropy": 0.0034450897946953773, "explorer_grad_norm": 0.38084179908037186, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.9749984741211, "explorer_subspace_norm": 121.4040756225586, "final_distance": 3.168457508087158, "max_param_abs": 2.159163475036621, "policy_loss": -0.001656242737077207, "value_loss": 0.12334729917347431}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:19:34.774071+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016374.774086, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.932483092390531, "elapsed_seconds": 14156.939132605912, "env_steps_per_second": 4096.930802394695, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.491631517803305, "eta_seconds": 5369.873464091898, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0005145524231314202, "chaser_grad_norm": 0.021116286981850863, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.860000610351562, "chaser_subspace_norm": 52.13620376586914, "collisions_per_episode": 1.125, "entropy": 0.002387709857430309, "explorer_grad_norm": 0.2525097243487835, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.125, "explorer_subspace_norm": 119.46451568603516, "final_distance": 3.4527671337127686, "max_param_abs": 2.180997610092163, "policy_loss": -0.00031888963405890536, "value_loss": 0.07078775204718113}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:27:49.117876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016869.117887, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.066998258098611, "elapsed_seconds": 14641.193729154998, "env_steps_per_second": 4098.026507259585, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3556660860328704, "eta_seconds": 4880.397909718334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005770839916344472, "chaser_grad_norm": 0.07790196500718594, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 51.55928421020508, "collisions_per_episode": 0.7250000238418579, "entropy": 0.004281870438717306, "explorer_grad_norm": 1.035013571381569, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 116.98993682861328, "final_distance": 3.6269309520721436, "max_param_abs": 2.206885576248169, "policy_loss": -0.0004720029999729647, "value_loss": 0.06566426623612642}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:35:53.372473+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017353.3724883, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.201983206083306, "elapsed_seconds": 15127.139541899902, "env_steps_per_second": 4098.593777645095, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.219930608217734, "eta_seconds": 4391.750189583842, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0011227785522094358, "chaser_grad_norm": 0.06945330277085304, "chaser_new_fields": 94.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 51.186004638671875, "collisions_per_episode": 1.2000000476837158, "entropy": 0.00486322829965502, "explorer_grad_norm": 0.4907003194093704, "explorer_new_fields": 97.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.01250457763672, "explorer_subspace_norm": 120.90361022949219, "final_distance": 3.6959824562072754, "max_param_abs": 2.233208656311035, "policy_loss": -0.000521478934562225, "value_loss": 0.09477586857974529}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:43:59.318285+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017839.3182957, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.341138527599138, "elapsed_seconds": 15628.098699356895, "env_steps_per_second": 4095.187855617628, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0852846318997844, "eta_seconds": 3907.0246748392237, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 4.187869177676301e-05, "chaser_grad_norm": 0.06548048881813884, "chaser_new_fields": 94.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 55.60856246948242, "collisions_per_episode": 0.949999988079071, "entropy": 0.0018750086892396212, "explorer_grad_norm": 0.3297066241502762, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 122.14292907714844, "final_distance": 3.239708423614502, "max_param_abs": 2.250119686126709, "policy_loss": -6.218791421552794e-05, "value_loss": 0.07839360274374485}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:20.277443+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018340.2774565, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.476195965257195, "elapsed_seconds": 16114.305474925903, "env_steps_per_second": 4095.739658323902, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9494961138424354, "eta_seconds": 3418.1860098327675, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0006283438149464853, "chaser_grad_norm": 0.019034171476960182, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 55.20565414428711, "collisions_per_episode": 1.0500000715255737, "entropy": 0.001432027667760849, "explorer_grad_norm": 0.4446190148591995, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 129.58328247070312, "final_distance": 3.658008575439453, "max_param_abs": 2.2777516841888428, "policy_loss": -0.0006528770187870236, "value_loss": 0.11454243957996368}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:26.484220+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018826.4842336, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.611858385852765, "elapsed_seconds": 16602.690189069952, "env_steps_per_second": 4095.721791205044, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8138573622093115, "eta_seconds": 2929.8865039535212, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00019735611764301695, "chaser_grad_norm": 0.014091947115957737, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.819999694824219, "chaser_subspace_norm": 56.74790954589844, "collisions_per_episode": 1.024999976158142, "entropy": 0.0021346345893107355, "explorer_grad_norm": 0.311550859361887, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 126.95128631591797, "final_distance": 3.3504257202148438, "max_param_abs": 2.2911319732666016, "policy_loss": -0.0001434955659351811, "value_loss": 0.07897400483489037}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:34.868934+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019314.8689485, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.748631615085275, "elapsed_seconds": 17095.07381430699, "env_steps_per_second": 4094.7468703771547, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6783759450121821, "eta_seconds": 2442.1534020438557, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 4.781807160392226e-05, "chaser_grad_norm": 0.035776169039309025, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.795000076293945, "chaser_subspace_norm": 57.92525100708008, "collisions_per_episode": 0.925000011920929, "entropy": 0.0018522321479395032, "explorer_grad_norm": 0.1872963383793831, "explorer_new_fields": 97.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 127.0145263671875, "final_distance": 3.466792345046997, "max_param_abs": 2.297224760055542, "policy_loss": -9.709990797279744e-05, "value_loss": 0.06055283918976784}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:16:47.252558+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019807.25257, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.885716232379426, "elapsed_seconds": 17588.578436565935, "env_steps_per_second": 4093.5656204207467, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5428573591532696, "eta_seconds": 1954.2864929517707, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0002409969932853917, "chaser_grad_norm": 0.0705314502120018, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 59.897953033447266, "collisions_per_episode": 1.100000023841858, "entropy": 0.002494490414392203, "explorer_grad_norm": 0.20680732280015945, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 129.95164489746094, "final_distance": 3.225666046142578, "max_param_abs": 2.3122336864471436, "policy_loss": -0.00024391081492591837, "value_loss": 0.1011340543627739}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:00.757180+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020300.757196, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.022272800527475, "elapsed_seconds": 18080.182081898907, "env_steps_per_second": 4092.879134999729, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40721130815087625, "eta_seconds": 1465.9607093431546, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0006484899224137308, "chaser_grad_norm": 0.08261850755661726, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.795001029968262, "chaser_subspace_norm": 59.100555419921875, "collisions_per_episode": 0.824999988079071, "entropy": 0.003048810991458595, "explorer_grad_norm": 0.3042645677924156, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 126.07718658447266, "final_distance": 3.82427978515625, "max_param_abs": 2.328068971633911, "policy_loss": -0.0004626124133366716, "value_loss": 0.0969211794435978}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:12.360826+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020792.3608377, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.16034138592416, "elapsed_seconds": 18577.228989326977, "env_steps_per_second": 4091.029940130665, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27159691504864003, "eta_seconds": 977.748894175104, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0005912727110923451, "chaser_grad_norm": 0.03518475824967027, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.850000381469727, "chaser_subspace_norm": 59.17195129394531, "collisions_per_episode": 1.25, "entropy": 0.0032758425804786384, "explorer_grad_norm": 0.9495732337236404, "explorer_new_fields": 97.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.67500305175781, "explorer_subspace_norm": 130.27586364746094, "final_distance": 3.5043532848358154, "max_param_abs": 2.353240966796875, "policy_loss": -0.0007552359548861354, "value_loss": 0.12067781016230583}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:29.407734+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021289.407749, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.297516541018591, "elapsed_seconds": 19071.059547666926, "env_steps_per_second": 4089.9667795092273, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13583375746201515, "eta_seconds": 489.0015268632545, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0012391397622986688, "chaser_grad_norm": 0.11135800927877426, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 53.68886184692383, "collisions_per_episode": 1.1750000715255737, "entropy": 0.003638271940872073, "explorer_grad_norm": 0.47454552724957466, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 126.65333557128906, "final_distance": 3.1356143951416016, "max_param_abs": 2.3651559352874756, "policy_loss": -0.0005628480403672143, "value_loss": 0.08898641727864742}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:43.238292+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021783.2383027, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.43102195080166, "elapsed_seconds": 19551.679022885975, "env_steps_per_second": 4091.7201999049284, "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-transformer/non_social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:43.857767+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022263.857778, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.431030371984137, "elapsed_seconds": 19551.709339142893, "env_steps_per_second": 4091.713855414088, "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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:43.888084+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022263.8880963, "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..2bde7ea15bf6d659bd0139796e6fb59d94cb26f9 --- /dev/null +++ b/non_social/seed_0002/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.431030371984137, + "elapsed_seconds": 19551.709339142893, + "env_steps_per_second": 4091.713855414088, + "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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 3.0426282704221208e-05, + "chaser_grad_norm": 0.03582112770527601, + "chaser_new_fields": 94.5250015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.820000648498535, + "chaser_subspace_norm": 58.25376892089844, + "collisions_per_episode": 0.9750000238418579, + "entropy": 0.0015975794522091746, + "explorer_grad_norm": 0.33543068170547485, + "explorer_new_fields": 97.5250015258789, + "explorer_partner_vision": 0.0, + "explorer_return": 95.57500457763672, + "explorer_subspace_norm": 127.81230163574219, + "final_distance": 3.7922744750976562, + "max_param_abs": 2.384674310684204, + "policy_loss": -4.3445531147767724e-05, + "value_loss": 0.07627437077462673 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:57:43.888084+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022263.8880963, + "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..13a7ef3070a8ec48b28b11fd2e2553653533900f --- /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:a06bb3351d710015bd4527f14150baa0ed7be3ea4e957deb0bac2ad812840a89 +size 29760352 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..47105c0883c343e96587d8febe5cea80363a6a24 --- /dev/null +++ b/non_social/seed_0003/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..ce2f33781498fd947512bf3caf69bcba0c417f70 --- /dev/null +++ b/non_social/seed_0003/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006961167227644991, "elapsed_seconds": 2.5060202019521967, "env_steps_per_second": 1596.1563266265728, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.921638338567217, "eta_seconds": 50117.89801884198, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005024423047657933, "chaser_grad_norm": 5.189185708761215, "chaser_new_fields": 31.17500114440918, "chaser_partner_vision": 0.0, "chaser_return": -3.8100006580352783, "chaser_subspace_norm": 0.029627107083797455, "collisions_per_episode": 1.725000023841858, "entropy": 2.755772113800049, "explorer_grad_norm": 9.743041276931763, "explorer_new_fields": 31.0, "explorer_partner_vision": 0.0, "explorer_return": -3.799999952316284, "explorer_subspace_norm": 0.038101453334093094, "final_distance": 4.984309673309326, "max_param_abs": 1.0012016296386719, "policy_loss": -0.01039676697925529, "value_loss": 7.190685272216797}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.674447+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.6744661, "update": 1} +{"device": "cuda", "elapsed_hours": 0.1340119316072216, "elapsed_seconds": 482.4429537859978, "env_steps_per_second": 4145.567852747955, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.226465332681643, "eta_seconds": 18815.275197653915, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.003495984075223113, "chaser_grad_norm": 0.4796079024672508, "chaser_new_fields": 89.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.780001163482666, "chaser_subspace_norm": 11.31277084350586, "collisions_per_episode": 1.649999976158142, "entropy": 0.12384101748466492, "explorer_grad_norm": 7.731855094432831, "explorer_new_fields": 88.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 81.8125, "explorer_subspace_norm": 14.753433227539062, "final_distance": 2.7396857738494873, "max_param_abs": 1.3429911136627197, "policy_loss": -0.001576730081808364, "value_loss": 0.7917164713144302}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:39:54.611379+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003194.611392, "update": 500} +{"device": "cuda", "elapsed_hours": 0.26942642090693375, "elapsed_seconds": 969.9351152649615, "env_steps_per_second": 4123.987199811095, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.119101997231741, "eta_seconds": 18428.76719003427, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0006954645487520228, "chaser_grad_norm": 0.3521268740296364, "chaser_new_fields": 89.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 7.770000457763672, "chaser_subspace_norm": 15.745429039001465, "collisions_per_episode": 3.625, "entropy": 0.03980019688606262, "explorer_grad_norm": 6.109767258167267, "explorer_new_fields": 90.25, "explorer_partner_vision": 0.0, "explorer_return": 84.70000457763672, "explorer_subspace_norm": 28.12783432006836, "final_distance": 2.597499132156372, "max_param_abs": 1.50963294506073, "policy_loss": -0.0020415620807305856, "value_loss": 0.4672740027308464}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:02.103538+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003682.1035526, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.40329279701804305, "elapsed_seconds": 1451.854069264955, "env_steps_per_second": 4132.646749433765, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.9739444965558635, "eta_seconds": 17906.20018760111, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004746300520819702, "chaser_grad_norm": 0.3817563150078058, "chaser_new_fields": 91.625, "chaser_partner_vision": 0.0, "chaser_return": 8.24000072479248, "chaser_subspace_norm": 20.85103988647461, "collisions_per_episode": 1.1750000715255737, "entropy": 0.026081882882863283, "explorer_grad_norm": 4.811295688152313, "explorer_new_fields": 93.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 89.3125, "explorer_subspace_norm": 39.89060592651367, "final_distance": 3.027148962020874, "max_param_abs": 1.6245406866073608, "policy_loss": -0.004421283919370467, "value_loss": 0.3692459166049957}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:04.022491+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004164.022504, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.538079933566106, "elapsed_seconds": 1937.0877608379815, "env_steps_per_second": 4129.910973439433, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.8427194020949536, "eta_seconds": 17433.789847541833, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0004433605326805701, "chaser_grad_norm": 0.05365113168954849, "chaser_new_fields": 92.125, "chaser_partner_vision": 0.0, "chaser_return": 8.34000015258789, "chaser_subspace_norm": 25.149837493896484, "collisions_per_episode": 0.925000011920929, "entropy": 0.014946554088965058, "explorer_grad_norm": 2.882696211338043, "explorer_new_fields": 94.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.11250305175781, "explorer_subspace_norm": 45.11609649658203, "final_distance": 2.9697210788726807, "max_param_abs": 1.7043564319610596, "policy_loss": -0.0007080689495202463, "value_loss": 0.14073776081204414}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:09.256186+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004649.2562013, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6727541947155583, "elapsed_seconds": 2421.91510097601, "env_steps_per_second": 4128.963891413902, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.709279363008908, "eta_seconds": 16953.40570683207, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0038921834673546485, "chaser_grad_norm": 0.15854962915182114, "chaser_new_fields": 91.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.225001335144043, "chaser_subspace_norm": 27.932138442993164, "collisions_per_episode": 1.25, "entropy": 0.01696440251544118, "explorer_grad_norm": 2.074641853570938, "explorer_new_fields": 94.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.73750305175781, "explorer_subspace_norm": 52.29011154174805, "final_distance": 3.3177506923675537, "max_param_abs": 1.7815338373184204, "policy_loss": -0.0045965834856218635, "value_loss": 0.1989050954580307}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:14.083525+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005134.0835383, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8094021273358117, "elapsed_seconds": 2913.847658408922, "env_steps_per_second": 4118.266088952805, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.586612054902933, "eta_seconds": 16511.803397650558, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.005292883627254774, "chaser_grad_norm": 0.08518719859421253, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.395000457763672, "chaser_subspace_norm": 30.823591232299805, "collisions_per_episode": 0.6000000238418579, "entropy": 0.011396759422495961, "explorer_grad_norm": 0.8767628371715546, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.75, "explorer_subspace_norm": 56.86723327636719, "final_distance": 3.3559136390686035, "max_param_abs": 1.8410812616348267, "policy_loss": -0.0024168349774527087, "value_loss": 0.15405790507793427}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:26.016083+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005626.0160983, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9458058168291528, "elapsed_seconds": 3404.90094058495, "env_steps_per_second": 4111.720206930557, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.458798850766007, "eta_seconds": 16051.675862757624, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0018205504827417118, "chaser_grad_norm": 0.0703201126307249, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.34000015258789, "chaser_subspace_norm": 32.318973541259766, "collisions_per_episode": 0.9750000238418579, "entropy": 0.013418101938441396, "explorer_grad_norm": 0.6616397947072983, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.7125015258789, "explorer_subspace_norm": 61.08270263671875, "final_distance": 3.17974853515625, "max_param_abs": 1.8905514478683472, "policy_loss": -0.0017446497985957876, "value_loss": 0.146530669182539}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:37.069364+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006117.0693789, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.080710129513043, "elapsed_seconds": 3890.556466246955, "env_steps_per_second": 4112.522241692197, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.322840518052172, "eta_seconds": 15562.22586498782, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0001880493554962115, "chaser_grad_norm": 0.0595186036080122, "chaser_new_fields": 92.25, "chaser_partner_vision": 0.0, "chaser_return": 8.370001792907715, "chaser_subspace_norm": 35.88082504272461, "collisions_per_episode": 0.9000000357627869, "entropy": 0.007024818332865834, "explorer_grad_norm": 0.5938415750861168, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 65.14606475830078, "final_distance": 3.142601728439331, "max_param_abs": 1.9357455968856812, "policy_loss": -0.000688961556879697, "value_loss": 0.10439493507146835}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:36:42.724890+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006602.724905, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.215635584145558, "elapsed_seconds": 4376.288102924009, "env_steps_per_second": 4113.074728323607, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.187189234279145, "eta_seconds": 15073.88124340492, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0002825130148482735, "chaser_grad_norm": 0.07820950355380774, "chaser_new_fields": 92.125, "chaser_partner_vision": 0.0, "chaser_return": 8.340001106262207, "chaser_subspace_norm": 39.256492614746094, "collisions_per_episode": 1.0500000715255737, "entropy": 0.006176907219924033, "explorer_grad_norm": 1.275507390499115, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.3375015258789, "explorer_subspace_norm": 66.62557220458984, "final_distance": 3.2689292430877686, "max_param_abs": 1.9760191440582275, "policy_loss": -0.0004644948476926203, "value_loss": 0.1531773805618286}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:44:48.456525+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007088.4565394, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.349486972211388, "elapsed_seconds": 4858.153099960997, "env_steps_per_second": 4116.790802694252, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.048460916634164, "eta_seconds": 14574.459299882988, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.006399905010243678, "chaser_grad_norm": 0.11409256421029568, "chaser_new_fields": 92.0, "chaser_partner_vision": 0.0, "chaser_return": 8.335000991821289, "chaser_subspace_norm": 39.89152908325195, "collisions_per_episode": 1.1750000715255737, "entropy": 0.011616653995588422, "explorer_grad_norm": 1.2792066633701324, "explorer_new_fields": 94.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.26250457763672, "explorer_subspace_norm": 65.928955078125, "final_distance": 3.333908796310425, "max_param_abs": 2.0132949352264404, "policy_loss": -0.0027241093716590115, "value_loss": 0.21737461909651756}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:52:50.321523+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007570.3215399, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4845045746044423, "elapsed_seconds": 5344.216468575993, "env_steps_per_second": 4116.599716602062, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9136938785026207, "eta_seconds": 14089.297962609435, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0012355989805410417, "chaser_grad_norm": 0.28674943186342716, "chaser_new_fields": 92.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.535000801086426, "chaser_subspace_norm": 41.7598762512207, "collisions_per_episode": 0.8500000238418579, "entropy": 0.006215645698830485, "explorer_grad_norm": 0.3715188130736351, "explorer_new_fields": 95.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 69.58002471923828, "final_distance": 3.4541115760803223, "max_param_abs": 2.048832654953003, "policy_loss": -0.0007717605554518059, "value_loss": 0.09684472344815731}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:00:56.384891+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008056.3849046, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6182306944688833, "elapsed_seconds": 5825.63050008798, "env_steps_per_second": 4119.725753227491, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7758716204273948, "eta_seconds": 13593.137833538622, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 4.920778848616081e-05, "chaser_grad_norm": 0.049862075597047806, "chaser_new_fields": 92.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.445000648498535, "chaser_subspace_norm": 41.50165939331055, "collisions_per_episode": 1.024999976158142, "entropy": 0.007846252177841961, "explorer_grad_norm": 1.0918368846178055, "explorer_new_fields": 94.5, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 76.08346557617188, "final_distance": 2.907108783721924, "max_param_abs": 2.0842835903167725, "policy_loss": -0.00036661602370813284, "value_loss": 0.1382632665336132}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:08:57.798923+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008537.7989385, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7589642375672054, "elapsed_seconds": 6332.27125524194, "env_steps_per_second": 4105.951711793276, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.653233416485734, "eta_seconds": 13151.640299348643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.002392652011237839, "chaser_grad_norm": 0.20674092881381512, "chaser_new_fields": 92.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.545001029968262, "chaser_subspace_norm": 42.17708969116211, "collisions_per_episode": 0.5, "entropy": 0.004808624275028706, "explorer_grad_norm": 0.7917798906564713, "explorer_new_fields": 95.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.4625015258789, "explorer_subspace_norm": 80.67359161376953, "final_distance": 3.240365743637085, "max_param_abs": 2.118196487426758, "policy_loss": -0.00039862397231615887, "value_loss": 0.0782720111310482}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:24.439677+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009044.4396899, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.892884754610261, "elapsed_seconds": 6814.385116596939, "env_steps_per_second": 4108.95473632741, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.515357401419056, "eta_seconds": 12655.286645108601, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0031535500349347823, "chaser_grad_norm": 0.13878342136740685, "chaser_new_fields": 92.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.505000114440918, "chaser_subspace_norm": 44.58604431152344, "collisions_per_episode": 0.925000011920929, "entropy": 0.013588826172053814, "explorer_grad_norm": 0.4408244490623474, "explorer_new_fields": 94.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 90.55000305175781, "explorer_subspace_norm": 78.97016906738281, "final_distance": 3.387521505355835, "max_param_abs": 2.1496760845184326, "policy_loss": -0.005201074440968512, "value_loss": 0.24234837293624878}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:26.553540+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009526.553556, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0274851760958295, "elapsed_seconds": 7298.946633944986, "env_steps_per_second": 4110.182126894849, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3791419601597155, "eta_seconds": 12164.911056574976, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00019590932616764256, "chaser_grad_norm": 0.03927200613543391, "chaser_new_fields": 93.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.555000305175781, "chaser_subspace_norm": 43.92483139038086, "collisions_per_episode": 0.9000000357627869, "entropy": 0.00862066401168704, "explorer_grad_norm": 0.3418860509991646, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 84.3548812866211, "final_distance": 3.2050235271453857, "max_param_abs": 2.185359477996826, "policy_loss": -0.0005563736307649947, "value_loss": 0.10596890561282635}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:31.115057+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010011.115069, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.161654925274165, "elapsed_seconds": 7781.957730986993, "env_steps_per_second": 4112.0757920052865, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.242482387911247, "eta_seconds": 11672.93659648049, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.008034418572132385, "chaser_grad_norm": 0.10141916945576668, "chaser_new_fields": 93.125, "chaser_partner_vision": 0.0, "chaser_return": 8.5600004196167, "chaser_subspace_norm": 45.13732147216797, "collisions_per_episode": 1.024999976158142, "entropy": 0.007907140767201781, "explorer_grad_norm": 0.6698158085346222, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 83.3248519897461, "final_distance": 3.2458255290985107, "max_param_abs": 2.2183961868286133, "policy_loss": -0.0011762413340576927, "value_loss": 0.12257060222327709}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:41:34.126155+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010494.1261694, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3011383428327585, "elapsed_seconds": 8284.09803419793, "env_steps_per_second": 4104.248870503848, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1133048167737316, "eta_seconds": 11207.897340385434, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 5.7506093533055136e-05, "chaser_grad_norm": 0.06035558506846428, "chaser_new_fields": 93.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.660000801086426, "chaser_subspace_norm": 48.30387496948242, "collisions_per_episode": 0.625, "entropy": 0.007350853877142072, "explorer_grad_norm": 1.0471926033496857, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 82.67190551757812, "final_distance": 3.156557083129883, "max_param_abs": 2.2406225204467773, "policy_loss": -0.00033004977623951604, "value_loss": 0.07093416340649128}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:49:56.266458+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010996.266472, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4384581666619245, "elapsed_seconds": 8778.449399982928, "env_steps_per_second": 4100.952042859643, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9803377592534632, "eta_seconds": 10729.215933312467, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0018212820362997206, "chaser_grad_norm": 0.08922404050827026, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 47.32369613647461, "collisions_per_episode": 0.925000011920929, "entropy": 0.00711581576615572, "explorer_grad_norm": 0.5106174945831299, "explorer_new_fields": 95.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.9124984741211, "explorer_subspace_norm": 87.91162109375, "final_distance": 2.5587334632873535, "max_param_abs": 2.281062602996826, "policy_loss": -0.0012821031226906143, "value_loss": 0.1541689820587635}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:10.617825+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011490.6178405, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5730465846694357, "elapsed_seconds": 9262.96770480997, "env_steps_per_second": 4102.356956320574, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8438935935820084, "eta_seconds": 10238.01693689523, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0018698370508686696, "chaser_grad_norm": 0.04480297304689884, "chaser_new_fields": 93.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 48.48903274536133, "collisions_per_episode": 0.8500000238418579, "entropy": 0.003263553255237639, "explorer_grad_norm": 0.9020937532186508, "explorer_new_fields": 95.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.4000015258789, "explorer_subspace_norm": 88.82749938964844, "final_distance": 2.9119434356689453, "max_param_abs": 2.3082261085510254, "policy_loss": -0.0006277953559781224, "value_loss": 0.06033980753272772}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:15.136128+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011975.1361413, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.708521510595003, "elapsed_seconds": 9750.677438142011, "env_steps_per_second": 4102.279072788402, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.708521510595003, "eta_seconds": 9750.677438142011, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0006473027324140901, "chaser_grad_norm": 0.10710760951042175, "chaser_new_fields": 93.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.55000114440918, "chaser_subspace_norm": 47.73188781738281, "collisions_per_episode": 1.100000023841858, "entropy": 0.006507194717414677, "explorer_grad_norm": 0.37044718861579895, "explorer_new_fields": 95.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 89.69586944580078, "final_distance": 2.9019782543182373, "max_param_abs": 2.330742120742798, "policy_loss": -0.0003978462544580452, "value_loss": 0.10910336673259735}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:22.845862+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012462.845876, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8425074297510826, "elapsed_seconds": 10233.026747103897, "env_steps_per_second": 4104.357492458098, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5717924364414557, "eta_seconds": 9258.45277118924, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0005266433975092966, "chaser_grad_norm": 0.06456576567143202, "chaser_new_fields": 93.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 51.4437255859375, "collisions_per_episode": 1.024999976158142, "entropy": 0.007705657742917538, "explorer_grad_norm": 0.4191744402050972, "explorer_new_fields": 95.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 90.49082946777344, "final_distance": 3.1663146018981934, "max_param_abs": 2.3666255474090576, "policy_loss": -0.0008286951470646287, "value_loss": 0.10708197765052319}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:22:25.195170+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012945.195182, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.97652851734139, "elapsed_seconds": 10715.502662429004, "env_steps_per_second": 4106.200276938387, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.43534151418841, "eta_seconds": 8767.229451078276, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0008063959727191428, "chaser_grad_norm": 0.045782639645040035, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.565000534057617, "chaser_subspace_norm": 51.45419692993164, "collisions_per_episode": 1.25, "entropy": 0.004481318756006658, "explorer_grad_norm": 0.3449392169713974, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 90.58958435058594, "final_distance": 3.308798313140869, "max_param_abs": 2.3922159671783447, "policy_loss": -0.0008177013732479921, "value_loss": 0.15009310096502304}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:30:27.671086+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013427.6711, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.111051448044988, "elapsed_seconds": 11199.785212961957, "env_steps_per_second": 4107.221623032768, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.299472809424556, "eta_seconds": 8278.102113928402, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0007127156194903783, "chaser_grad_norm": 0.03489385172724724, "chaser_new_fields": 93.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.565000534057617, "chaser_subspace_norm": 52.208133697509766, "collisions_per_episode": 1.024999976158142, "entropy": 0.0037381406873464584, "explorer_grad_norm": 0.4758947938680649, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 99.85450744628906, "final_distance": 2.635690212249756, "max_param_abs": 2.4100501537323, "policy_loss": -0.0008028766119352326, "value_loss": 0.11823840066790581}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:38:31.953636+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013911.9536479, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.245491730141641, "elapsed_seconds": 11683.770228509908, "env_steps_per_second": 4108.262920377688, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.163661153427761, "eta_seconds": 7789.180152339939, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0020081153855908696, "chaser_grad_norm": 0.04150939965620637, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.570000648498535, "chaser_subspace_norm": 53.23170852661133, "collisions_per_episode": 0.875, "entropy": 0.004793138708919287, "explorer_grad_norm": 0.7487403303384781, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 96.56257629394531, "final_distance": 3.115834951400757, "max_param_abs": 2.4359054565429688, "policy_loss": -0.000987323350628344, "value_loss": 0.08897869288921356}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:46:35.938653+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014395.9386683, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3880492816511025, "elapsed_seconds": 12196.977413943969, "env_steps_per_second": 4099.376288328486, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0328295689906613, "eta_seconds": 7318.18644836638, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005557573119234771, "chaser_grad_norm": 0.04495162144303322, "chaser_new_fields": 93.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.5600004196167, "chaser_subspace_norm": 55.89772415161133, "collisions_per_episode": 1.024999976158142, "entropy": 0.004529337864369154, "explorer_grad_norm": 0.43399157002568245, "explorer_new_fields": 95.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 94.74485778808594, "final_distance": 3.02925705909729, "max_param_abs": 2.445003032684326, "policy_loss": -0.00028155376063576654, "value_loss": 0.11668848805129528}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:09.145836+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014909.1458454, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5251567873161145, "elapsed_seconds": 12690.564434338012, "env_steps_per_second": 4097.5324832124, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8981613470163694, "eta_seconds": 6833.38084925893, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.004025487210471246, "chaser_grad_norm": 0.11131552793085575, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 52.96380615234375, "collisions_per_episode": 1.0500000715255737, "entropy": 0.005851391702890396, "explorer_grad_norm": 0.4314912408590317, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 93.7562484741211, "final_distance": 2.8668484687805176, "max_param_abs": 2.472851276397705, "policy_loss": -0.0010405651386875547, "value_loss": 0.12776139751076698}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:22.732860+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015402.7328765, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.662232038019445, "elapsed_seconds": 13184.035336870002, "env_steps_per_second": 4095.8628083304307, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7632969071945475, "eta_seconds": 6347.868865900371, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00040896350521180125, "chaser_grad_norm": 0.05079989228397608, "chaser_new_fields": 93.625, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 57.54914093017578, "collisions_per_episode": 1.125, "entropy": 0.005252829403616488, "explorer_grad_norm": 0.49959227815270424, "explorer_new_fields": 95.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.36250305175781, "explorer_subspace_norm": 98.56642150878906, "final_distance": 3.139416217803955, "max_param_abs": 2.5008347034454346, "policy_loss": -0.0006854486721579178, "value_loss": 0.1017832774668932}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:36.203761+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015896.203774, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.798527904025266, "elapsed_seconds": 13674.700454490958, "env_steps_per_second": 4095.153688109405, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6279405302965426, "eta_seconds": 5860.585909067553, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0022012519165470756, "chaser_grad_norm": 0.05405232170596719, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 59.76309585571289, "collisions_per_episode": 1.149999976158142, "entropy": 0.00722871208563447, "explorer_grad_norm": 0.6766604036092758, "explorer_new_fields": 95.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 104.74185943603516, "final_distance": 2.8553645610809326, "max_param_abs": 2.524658679962158, "policy_loss": -0.00035590077948199905, "value_loss": 0.12009123712778091}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:19:46.868878+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016386.8688936, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.937308964141929, "elapsed_seconds": 14174.312270910945, "env_steps_per_second": 4091.9092857175, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4934620208814213, "eta_seconds": 5376.463275173117, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.004644796858217504, "chaser_grad_norm": 0.4359093941748142, "chaser_new_fields": 93.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.670001029968262, "chaser_subspace_norm": 60.4807243347168, "collisions_per_episode": 0.9750000238418579, "entropy": 0.007179249660111964, "explorer_grad_norm": 0.3641759119927883, "explorer_new_fields": 95.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0625, "explorer_subspace_norm": 100.59651947021484, "final_distance": 2.7122199535369873, "max_param_abs": 2.5459628105163574, "policy_loss": -0.0011473420108529808, "value_loss": 0.12315200828015804}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:06.480695+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016886.480711, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.074349657038596, "elapsed_seconds": 14667.658765338943, "env_steps_per_second": 4090.6323878890366, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3581165523461984, "eta_seconds": 4889.219588446314, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0003340514101120573, "chaser_grad_norm": 0.07047029910609126, "chaser_new_fields": 93.75, "chaser_partner_vision": 0.0, "chaser_return": 8.695000648498535, "chaser_subspace_norm": 61.55753707885742, "collisions_per_episode": 0.925000011920929, "entropy": 0.0027071837102994323, "explorer_grad_norm": 0.37893088161945343, "explorer_new_fields": 95.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 107.49569702148438, "final_distance": 2.6742610931396484, "max_param_abs": 2.5662577152252197, "policy_loss": -0.00021887112181873114, "value_loss": 0.12173754349350929}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:19.827188+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017379.8272038, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.209102502566368, "elapsed_seconds": 15152.769009238924, "env_steps_per_second": 4091.661396157854, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2219975007450745, "eta_seconds": 4399.191002682268, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0005762351906208885, "chaser_grad_norm": 0.07829482853412628, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 67.62864685058594, "collisions_per_episode": 0.875, "entropy": 0.004984446335583925, "explorer_grad_norm": 0.22619431093335152, "explorer_new_fields": 95.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.4000015258789, "explorer_subspace_norm": 106.18888092041016, "final_distance": 2.723506450653076, "max_param_abs": 2.5889992713928223, "policy_loss": -0.00048407367150815617, "value_loss": 0.07387478835880756}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:24.937431+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017864.9374437, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.343068402223031, "elapsed_seconds": 15635.04624800291, "env_steps_per_second": 4093.3681285512557, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0857671005557576, "eta_seconds": 3908.7615620007273, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005539466711863561, "chaser_grad_norm": 0.06942280102521181, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 62.85855484008789, "collisions_per_episode": 0.875, "entropy": 0.005629157414659858, "explorer_grad_norm": 0.31974368914961815, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 110.23323822021484, "final_distance": 3.025721311569214, "max_param_abs": 2.6085193157196045, "policy_loss": -0.0007279603790912859, "value_loss": 0.1100351344794035}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:27.214672+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018347.2146873, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.476395920648033, "elapsed_seconds": 16115.025314332917, "env_steps_per_second": 4095.5567064048437, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9495385286223099, "eta_seconds": 3418.3387030403155, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0003900722442824467, "chaser_grad_norm": 0.13584143295884132, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 66.25489807128906, "collisions_per_episode": 1.0, "entropy": 0.005487023736350238, "explorer_grad_norm": 0.4194389581680298, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 110.97061920166016, "final_distance": 2.7326693534851074, "max_param_abs": 2.6253645420074463, "policy_loss": -0.000591823476483011, "value_loss": 0.12209559418261051}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:27.193738+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018827.1937509, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.612801431243586, "elapsed_seconds": 16606.08515247691, "env_steps_per_second": 4094.8844580540613, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8140237819841623, "eta_seconds": 2930.485615142984, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0004943212216756976, "chaser_grad_norm": 0.04927615146152675, "chaser_new_fields": 93.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.705000877380371, "chaser_subspace_norm": 66.02193450927734, "collisions_per_episode": 0.7250000238418579, "entropy": 0.004525314900092781, "explorer_grad_norm": 0.6757950857281685, "explorer_new_fields": 95.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.4000015258789, "explorer_subspace_norm": 112.50408172607422, "final_distance": 2.9537694454193115, "max_param_abs": 2.631089925765991, "policy_loss": -0.0002603902962299287, "value_loss": 0.06927836500108242}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:38.253576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019318.2535892, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.748318701824141, "elapsed_seconds": 17093.947326566908, "env_steps_per_second": 4095.016713384162, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6783312431177344, "eta_seconds": 2441.992475223844, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0011644642110459902, "chaser_grad_norm": 0.03152832528576255, "chaser_new_fields": 93.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.574999809265137, "chaser_subspace_norm": 71.01042175292969, "collisions_per_episode": 1.1750000715255737, "entropy": 0.005007880623452365, "explorer_grad_norm": 0.4314521551132202, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 111.36642456054688, "final_distance": 3.0410330295562744, "max_param_abs": 2.6621429920196533, "policy_loss": -0.0012926242689133005, "value_loss": 0.14316125959157944}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:16:46.115748+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019806.1157608, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.8837019592194375, "elapsed_seconds": 17581.327053189976, "env_steps_per_second": 4095.2540034192834, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5426335510243819, "eta_seconds": 1953.480783687775, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0005835298663985355, "chaser_grad_norm": 0.07682760432362556, "chaser_new_fields": 93.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.699999809265137, "chaser_subspace_norm": 75.85527801513672, "collisions_per_episode": 0.925000011920929, "entropy": 0.0035212296061217785, "explorer_grad_norm": 0.4473385736346245, "explorer_new_fields": 95.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.25, "explorer_subspace_norm": 113.15235900878906, "final_distance": 2.944685697555542, "max_param_abs": 2.6787211894989014, "policy_loss": -0.0009910173625691243, "value_loss": 0.08652106299996376}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:24:53.495476+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020293.495492, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.018611647756382, "elapsed_seconds": 18067.001931922976, "env_steps_per_second": 4095.864951962384, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40691445792619313, "eta_seconds": 1464.8920485342953, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 4.9106388614772456e-05, "chaser_grad_norm": 0.03466647490859032, "chaser_new_fields": 93.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.65999984741211, "chaser_subspace_norm": 75.69221496582031, "collisions_per_episode": 0.949999988079071, "entropy": 0.003670561476610601, "explorer_grad_norm": 0.2681044861674309, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.2874984741211, "explorer_subspace_norm": 118.05979919433594, "final_distance": 2.848959445953369, "max_param_abs": 2.7065818309783936, "policy_loss": -0.00013019449774542835, "value_loss": 0.11256165988743305}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:32:59.170354+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020779.1703684, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.152573043844152, "elapsed_seconds": 18549.262957838946, "env_steps_per_second": 4097.197833290853, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2711880549391658, "eta_seconds": 976.276997780997, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 7.057428441470392e-05, "chaser_grad_norm": 0.03550256649032235, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 77.3493423461914, "collisions_per_episode": 1.0750000476837158, "entropy": 0.003740044543519616, "explorer_grad_norm": 0.3448968231678009, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.7249984741211, "explorer_subspace_norm": 115.06394958496094, "final_distance": 2.810687780380249, "max_param_abs": 2.711897850036621, "policy_loss": -0.00033303190628153345, "value_loss": 0.12280823849141598}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:01.431381+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021261.4313953, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.289410399953582, "elapsed_seconds": 19041.877439832897, "env_steps_per_second": 4096.234746098886, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1356259076911175, "eta_seconds": 488.25326768802296, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.001099711209915366, "chaser_grad_norm": 0.020938430912792683, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.660000801086426, "chaser_subspace_norm": 81.31414794921875, "collisions_per_episode": 1.0, "entropy": 0.00502233556471765, "explorer_grad_norm": 0.39622850716114044, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 119.11949157714844, "final_distance": 3.1853957176208496, "max_param_abs": 2.7196757793426514, "policy_loss": -0.0005644467060932357, "value_loss": 0.12597087770700455}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:14.045863+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021754.0458794, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.426759173181652, "elapsed_seconds": 19536.333023453946, "env_steps_per_second": 4094.9342900716133, "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-transformer/non_social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:28.501446+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022248.5014582, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.427600972299988, "elapsed_seconds": 19539.363500279956, "env_steps_per_second": 4094.2991822048743, "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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:31.531925+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022251.531941, "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..03613a6b8df84a9a61f8ac37c122b996711ef17c --- /dev/null +++ b/non_social/seed_0003/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.427600972299988, + "elapsed_seconds": 19539.363500279956, + "env_steps_per_second": 4094.2991822048743, + "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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0002613905893603663, + "chaser_grad_norm": 0.02223971695639193, + "chaser_new_fields": 93.7249984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 8.6850004196167, + "chaser_subspace_norm": 82.19728088378906, + "collisions_per_episode": 1.0, + "entropy": 0.0022482319036498666, + "explorer_grad_norm": 0.44222569093108177, + "explorer_new_fields": 95.7750015258789, + "explorer_partner_vision": 0.0, + "explorer_return": 93.0999984741211, + "explorer_subspace_norm": 121.7215576171875, + "final_distance": 2.864136219024658, + "max_param_abs": 2.740300178527832, + "policy_loss": -0.00019148159428716838, + "value_loss": 0.08482817187905312 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:57:31.531925+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022251.531941, + "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..ad2e5fee8ce0e497dedfb9ecc0bc96f4c8f03ae1 --- /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 sha256:7fc0051ad27e22e8e395faf9f846864d859748404592f14a524f3596f1f93559 +size 29760360 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..1b8dd2dc7db9ff6712fcff9962a68d613d18ea06 --- /dev/null +++ b/non_social/seed_0004/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..29e582e36811eac28762cadb70badec002df09b2 --- /dev/null +++ b/non_social/seed_0004/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007179040744409172, "elapsed_seconds": 2.584454667987302, "env_steps_per_second": 1547.7152876955213, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.357363584743903, "eta_seconds": 51686.50890507805, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004168721541645493, "chaser_grad_norm": 6.509516932070255, "chaser_new_fields": 34.95000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.06000018119812, "chaser_subspace_norm": 0.032386090606451035, "collisions_per_episode": 2.049999952316284, "entropy": 2.7535924315452576, "explorer_grad_norm": 6.642140626907349, "explorer_new_fields": 31.42500114440918, "explorer_partner_vision": 0.0, "explorer_return": -3.3125, "explorer_subspace_norm": 0.03992241621017456, "final_distance": 5.027736663818359, "max_param_abs": 1.0011932849884033, "policy_loss": -0.013245799133557057, "value_loss": 7.792621850967407}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.576904+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.576917, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13611594677527641, "elapsed_seconds": 490.0174083909951, "env_steps_per_second": 4081.4876487085094, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.30852192423578, "eta_seconds": 19110.67892724881, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.001165815932237635, "chaser_grad_norm": 0.1750371716916561, "chaser_new_fields": 92.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.480000495910645, "chaser_subspace_norm": 13.320653915405273, "collisions_per_episode": 0.550000011920929, "entropy": 0.1037871316075325, "explorer_grad_norm": 8.205935001373291, "explorer_new_fields": 87.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 81.2874984741211, "explorer_subspace_norm": 15.509645462036133, "final_distance": 3.4141125679016113, "max_param_abs": 1.3126949071884155, "policy_loss": -0.0033158181114174567, "value_loss": 0.5582300126552582}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:02.009857+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003202.0098708, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27003779055110905, "elapsed_seconds": 972.1360459839925, "env_steps_per_second": 4114.65043038417, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.130718020471072, "eta_seconds": 18470.584873695858, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0019922951621613727, "chaser_grad_norm": 0.31193093210458755, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.760000228881836, "chaser_subspace_norm": 19.159109115600586, "collisions_per_episode": 0.22500000894069672, "entropy": 0.02209476660937071, "explorer_grad_norm": 4.885534763336182, "explorer_new_fields": 93.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.0999984741211, "explorer_subspace_norm": 29.250194549560547, "final_distance": 4.79291296005249, "max_param_abs": 1.4873065948486328, "policy_loss": -0.00244688695413231, "value_loss": 0.11829400807619095}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:04.128494+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003684.1285083, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4079382929050027, "elapsed_seconds": 1468.5778544580098, "env_steps_per_second": 4085.585235938579, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.031238945828368, "eta_seconds": 18112.460204982122, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0036912858272062365, "chaser_grad_norm": 0.6820209696888924, "chaser_new_fields": 87.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.540000915527344, "chaser_subspace_norm": 25.168725967407227, "collisions_per_episode": 0.7750000357627869, "entropy": 0.017583402805030346, "explorer_grad_norm": 3.2422459721565247, "explorer_new_fields": 93.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 89.8375015258789, "explorer_subspace_norm": 37.35585021972656, "final_distance": 3.5126266479492188, "max_param_abs": 1.6164562702178955, "policy_loss": -0.0008014342512403516, "value_loss": 0.20651323720812798}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:20.570304+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004180.5703175, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5459863309944436, "elapsed_seconds": 1965.5507915799972, "env_steps_per_second": 4070.1059643283215, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.913876978949993, "eta_seconds": 17689.957124219975, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.001175347762921497, "chaser_grad_norm": 0.24416987970471382, "chaser_new_fields": 89.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.850000858306885, "chaser_subspace_norm": 30.818607330322266, "collisions_per_episode": 0.675000011920929, "entropy": 0.013233349192887545, "explorer_grad_norm": 1.5040321201086044, "explorer_new_fields": 94.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.8499984741211, "explorer_subspace_norm": 45.67414855957031, "final_distance": 3.4709484577178955, "max_param_abs": 1.7207281589508057, "policy_loss": -0.00155099803162706, "value_loss": 0.11385388672351837}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:37.543241+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004677.5432556, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.682896484242778, "elapsed_seconds": 2458.427343274001, "env_steps_per_second": 4067.6410581581554, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.7802753896994465, "eta_seconds": 17208.991402918007, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": -7.057968711263879e-05, "chaser_grad_norm": 0.068180195055902, "chaser_new_fields": 88.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.595001220703125, "chaser_subspace_norm": 33.693416595458984, "collisions_per_episode": 0.699999988079071, "entropy": 0.006714100483804941, "explorer_grad_norm": 1.265616461634636, "explorer_new_fields": 94.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.63750457763672, "explorer_subspace_norm": 54.46622848510742, "final_distance": 4.577891826629639, "max_param_abs": 1.8177306652069092, "policy_loss": -1.9406676939226486e-05, "value_loss": 0.09330222010612488}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:50.419790+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005170.4198022, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8202148621105718, "elapsed_seconds": 2952.7735035980586, "env_steps_per_second": 4063.9757791708635, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.647884218626573, "eta_seconds": 16732.383187055664, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004725548817338096, "chaser_grad_norm": 0.2967050150036812, "chaser_new_fields": 88.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.599999904632568, "chaser_subspace_norm": 37.740169525146484, "collisions_per_episode": 3.075000047683716, "entropy": 0.011035462841391563, "explorer_grad_norm": 0.8352572470903397, "explorer_new_fields": 92.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 87.88750457763672, "explorer_subspace_norm": 57.533329010009766, "final_distance": 4.167410373687744, "max_param_abs": 1.8849718570709229, "policy_loss": 0.0017718010931275785, "value_loss": 0.351640447974205}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:21:04.765951+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005664.765965, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9561326472975068, "elapsed_seconds": 3442.0775302710244, "env_steps_per_second": 4067.3110576035338, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.507482480116818, "eta_seconds": 16226.936928420544, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005448501374183401, "chaser_grad_norm": 0.06794039160013199, "chaser_new_fields": 91.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.305000305175781, "chaser_subspace_norm": 39.59990692138672, "collisions_per_episode": 0.699999988079071, "entropy": 0.006905361427925527, "explorer_grad_norm": 0.8802538886666298, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 65.91787719726562, "final_distance": 4.052903175354004, "max_param_abs": 1.9597896337509155, "policy_loss": -0.00037312930973776304, "value_loss": 0.10577482357621193}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:14.069979+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006154.069993, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0919844198711022, "elapsed_seconds": 3931.143911535968, "env_steps_per_second": 4070.062139686083, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.367937679484409, "eta_seconds": 15724.575646143872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0007529899503896237, "chaser_grad_norm": 0.05311888735741377, "chaser_new_fields": 91.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.215001106262207, "chaser_subspace_norm": 42.80828094482422, "collisions_per_episode": 0.875, "entropy": 0.008078583749011159, "explorer_grad_norm": 0.58997692912817, "explorer_new_fields": 94.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.48750305175781, "explorer_subspace_norm": 70.58789825439453, "final_distance": 4.197776794433594, "max_param_abs": 2.0185790061950684, "policy_loss": -0.000664450679323636, "value_loss": 0.13967562094330788}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:23.136360+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006643.136374, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2263701481272196, "elapsed_seconds": 4414.93253325799, "env_steps_per_second": 4077.07249531103, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.224163843549312, "eta_seconds": 15206.989836777522, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00033688677544142787, "chaser_grad_norm": 0.046388428658246994, "chaser_new_fields": 91.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.17500114440918, "chaser_subspace_norm": 45.70652389526367, "collisions_per_episode": 1.0, "entropy": 0.008538259658962488, "explorer_grad_norm": 0.6624040454626083, "explorer_new_fields": 94.25, "explorer_partner_vision": 0.0, "explorer_return": 91.07500457763672, "explorer_subspace_norm": 73.56023406982422, "final_distance": 4.0290141105651855, "max_param_abs": 2.0689001083374023, "policy_loss": -0.0007023077382442011, "value_loss": 0.17052556574344635}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:26.924980+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007126.9249928, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3600353067249913, "elapsed_seconds": 4896.127104209969, "env_steps_per_second": 4084.861273883773, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.080105920174974, "eta_seconds": 14688.381312629906, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00607271511547558, "chaser_grad_norm": 0.15301267057657242, "chaser_new_fields": 91.375, "chaser_partner_vision": 0.0, "chaser_return": 8.210000991821289, "chaser_subspace_norm": 47.38425827026367, "collisions_per_episode": 0.7250000238418579, "entropy": 0.010898770997300744, "explorer_grad_norm": 0.9010711759328842, "explorer_new_fields": 94.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 90.92500305175781, "explorer_subspace_norm": 78.74980163574219, "final_distance": 3.634722948074341, "max_param_abs": 2.1259748935699463, "policy_loss": -0.0020766167314718587, "value_loss": 0.18118872493505478}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:28.119552+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007608.119565, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4964520221661142, "elapsed_seconds": 5387.227279798011, "env_steps_per_second": 4083.7334044731206, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9451916948015735, "eta_seconds": 14202.690101285665, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.001186162677462077, "chaser_grad_norm": 0.07213358487933874, "chaser_new_fields": 91.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.160000801086426, "chaser_subspace_norm": 53.68938446044922, "collisions_per_episode": 1.4250000715255737, "entropy": 0.007026613922789693, "explorer_grad_norm": 0.4948671758174896, "explorer_new_fields": 93.75, "explorer_partner_vision": 0.0, "explorer_return": 90.17500305175781, "explorer_subspace_norm": 82.62323760986328, "final_distance": 3.7338340282440186, "max_param_abs": 2.1736793518066406, "policy_loss": -0.0012927470467936075, "value_loss": 0.2303939014673233}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:39.219728+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008099.2197392, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6302905221366544, "elapsed_seconds": 5869.045879691956, "env_steps_per_second": 4089.25070479423, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.80401121831886, "eta_seconds": 13694.440385947895, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0005560482332418459, "chaser_grad_norm": 0.06472470052540302, "chaser_new_fields": 91.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.265000343322754, "chaser_subspace_norm": 56.36345672607422, "collisions_per_episode": 0.824999988079071, "entropy": 0.006473298300988972, "explorer_grad_norm": 0.36392417550086975, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.7874984741211, "explorer_subspace_norm": 84.27880859375, "final_distance": 3.650900363922119, "max_param_abs": 2.226118564605713, "policy_loss": -0.0016544239111198067, "value_loss": 0.14978458732366562}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:41.038327+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008581.038341, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7643612070447288, "elapsed_seconds": 6351.700345361023, "env_steps_per_second": 4093.392097596221, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6644425069390523, "eta_seconds": 13191.993024980588, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.00018242453012728976, "chaser_grad_norm": 0.06973688397556543, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 55.79123306274414, "collisions_per_episode": 0.625, "entropy": 0.004662334104068577, "explorer_grad_norm": 0.2566753923892975, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 86.42034149169922, "final_distance": 3.713762044906616, "max_param_abs": 2.2646422386169434, "policy_loss": -8.36074383117058e-05, "value_loss": 0.076368048787117}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:43.692791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009063.6928043, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8983404127383254, "elapsed_seconds": 6834.025485857972, "env_steps_per_second": 4097.1459731811, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5254893379426044, "eta_seconds": 12691.761616593376, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004217445315669366, "chaser_grad_norm": 0.26001538522541523, "chaser_new_fields": 91.75, "chaser_partner_vision": 0.0, "chaser_return": 8.265000343322754, "chaser_subspace_norm": 55.53413391113281, "collisions_per_episode": 0.925000011920929, "entropy": 0.006662545842118561, "explorer_grad_norm": 0.5009182542562485, "explorer_new_fields": 94.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 90.10924530029297, "final_distance": 3.585752248764038, "max_param_abs": 2.3055925369262695, "policy_loss": -0.0016401482354879704, "value_loss": 0.11595870740711689}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:46.017934+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009546.0179505, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.034343741137224, "elapsed_seconds": 7323.637468094006, "env_steps_per_second": 4096.3251022046525, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.390572901895373, "eta_seconds": 12206.062446823344, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.001021832967690317, "chaser_grad_norm": 0.12410663813352585, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.385001182556152, "chaser_subspace_norm": 58.660423278808594, "collisions_per_episode": 0.75, "entropy": 0.0037105209194123745, "explorer_grad_norm": 0.5163974612951279, "explorer_new_fields": 94.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.20000457763672, "explorer_subspace_norm": 94.83783721923828, "final_distance": 4.33816385269165, "max_param_abs": 2.3477609157562256, "policy_loss": -0.00017890909968798496, "value_loss": 0.12360158190131187}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:55.629915+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010035.6299274, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.170913896232507, "elapsed_seconds": 7815.2900264370255, "env_steps_per_second": 4094.537744824901, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2563708443487607, "eta_seconds": 11722.935039655538, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005332072360353379, "chaser_grad_norm": 0.07196172885596752, "chaser_new_fields": 92.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.410000801086426, "chaser_subspace_norm": 59.72168731689453, "collisions_per_episode": 0.6000000238418579, "entropy": 0.004704791703261435, "explorer_grad_norm": 0.3538786545395851, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 100.75301361083984, "final_distance": 3.4285435676574707, "max_param_abs": 2.377171277999878, "policy_loss": -0.0005098494933299946, "value_loss": 0.083933986723423}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:07.282474+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010527.282488, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3048166720188843, "elapsed_seconds": 8297.340019267984, "env_steps_per_second": 4097.698771057424, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1182813797902553, "eta_seconds": 11225.81296724492, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0018567237811510928, "chaser_grad_norm": 0.2961665950715542, "chaser_new_fields": 92.0, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 61.537147521972656, "collisions_per_episode": 0.45000001788139343, "entropy": 0.007917658076621592, "explorer_grad_norm": 0.24600420892238617, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 102.79935455322266, "final_distance": 3.600990056991577, "max_param_abs": 2.410496950149536, "policy_loss": -0.00022455692280232142, "value_loss": 0.047318145632743835}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:09.332466+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011009.3324792, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.444835862054994, "elapsed_seconds": 8801.409103397978, "env_steps_per_second": 4090.2541373859563, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9881327202894368, "eta_seconds": 10757.277793041972, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.01082343613169795, "chaser_grad_norm": 0.07303811423480511, "chaser_new_fields": 92.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.350001335144043, "chaser_subspace_norm": 64.40133666992188, "collisions_per_episode": 1.125, "entropy": 0.005557489930652082, "explorer_grad_norm": 0.3703003525733948, "explorer_new_fields": 93.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 89.9124984741211, "explorer_subspace_norm": 101.63013458251953, "final_distance": 3.673555374145508, "max_param_abs": 2.434584379196167, "policy_loss": -0.0057121851332482265, "value_loss": 0.17832371965050697}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:33.401552+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011513.4015658, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.581718608397512, "elapsed_seconds": 9294.186990231043, "env_steps_per_second": 4088.5770901684177, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8534784619130393, "eta_seconds": 10272.522462886942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.01158037428318126, "chaser_grad_norm": 0.42690640687942505, "chaser_new_fields": 89.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.835000514984131, "chaser_subspace_norm": 64.35760498046875, "collisions_per_episode": 3.1500000953674316, "entropy": 0.009429939789697528, "explorer_grad_norm": 0.4501849487423897, "explorer_new_fields": 92.75, "explorer_partner_vision": 0.0, "explorer_return": 88.9000015258789, "explorer_subspace_norm": 105.53968811035156, "final_distance": 3.7971291542053223, "max_param_abs": 2.472315549850464, "policy_loss": -0.010460890986955462, "value_loss": 0.36160628497600555}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:46.179438+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012006.1794524, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7168194757572364, "elapsed_seconds": 9780.55011272605, "env_steps_per_second": 4089.749506825147, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7168194757572364, "eta_seconds": 9780.55011272605, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.000252617621805222, "chaser_grad_norm": 0.02687776880338788, "chaser_new_fields": 92.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.51500129699707, "chaser_subspace_norm": 65.24219512939453, "collisions_per_episode": 0.4749999940395355, "entropy": 0.002895449462812394, "explorer_grad_norm": 0.38908619061112404, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 106.086669921875, "final_distance": 3.815507650375366, "max_param_abs": 2.503999948501587, "policy_loss": -2.349901598819315e-05, "value_loss": 0.0640158373862505}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:52.542562+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012492.5425782, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8534589431105672, "elapsed_seconds": 10272.452195198042, "env_steps_per_second": 4088.605057673893, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.581700948528608, "eta_seconds": 9294.12341470299, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -0.0002584837099098153, "chaser_grad_norm": 0.07851303089410067, "chaser_new_fields": 92.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.530000686645508, "chaser_subspace_norm": 68.64289855957031, "collisions_per_episode": 0.5250000357627869, "entropy": 0.002932213246822357, "explorer_grad_norm": 0.16077022813260555, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 108.86304473876953, "final_distance": 3.8342368602752686, "max_param_abs": 2.519569158554077, "policy_loss": -0.0005988772140135268, "value_loss": 0.06520954519510269}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:04.444644+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012984.4446592, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.992207642904153, "elapsed_seconds": 10771.94751445495, "env_steps_per_second": 4084.683845790754, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4481698896488524, "eta_seconds": 8813.411602735869, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0028229508196702113, "chaser_grad_norm": 0.338572159409523, "chaser_new_fields": 92.625, "chaser_partner_vision": 0.0, "chaser_return": 8.470000267028809, "chaser_subspace_norm": 65.50027465820312, "collisions_per_episode": 0.625, "entropy": 0.005788428126834333, "explorer_grad_norm": 0.46209583431482315, "explorer_new_fields": 94.875, "explorer_partner_vision": 0.0, "explorer_return": 91.9375, "explorer_subspace_norm": 112.18248748779297, "final_distance": 3.9190170764923096, "max_param_abs": 2.553093671798706, "policy_loss": -0.0014898449838938177, "value_loss": 0.08572590164840221}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:23.939964+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013483.9399781, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1285971167527733, "elapsed_seconds": 11262.949620309984, "env_steps_per_second": 4084.187672920974, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3124413471650938, "eta_seconds": 8324.788849794337, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00043558126886455995, "chaser_grad_norm": 0.053662254475057125, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 70.17605590820312, "collisions_per_episode": 0.32500001788139343, "entropy": 0.0020849284483119845, "explorer_grad_norm": 0.3590114116668701, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 116.16581726074219, "final_distance": 4.035409450531006, "max_param_abs": 2.578110933303833, "policy_loss": -0.0006839056420631096, "value_loss": 0.02623558323830366}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:34.942069+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013974.9420834, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.267499841000568, "elapsed_seconds": 11762.999427602044, "env_steps_per_second": 4080.5918843596405, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.178333227333712, "eta_seconds": 7841.9996184013635, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0020530864643185653, "chaser_grad_norm": 0.09303068090230227, "chaser_new_fields": 92.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.375000953674316, "chaser_subspace_norm": 71.1871337890625, "collisions_per_episode": 0.675000011920929, "entropy": 0.0027941627195104957, "explorer_grad_norm": 0.37502408772706985, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 121.38587951660156, "final_distance": 3.820488691329956, "max_param_abs": 2.6030306816101074, "policy_loss": -0.0004895105308387881, "value_loss": 0.07220060378313065}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:54.991876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014474.9918914, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.401548980147765, "elapsed_seconds": 12245.576328531955, "env_steps_per_second": 4083.1071285310572, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.040929388088659, "eta_seconds": 7347.345797119174, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00044117828131540193, "chaser_grad_norm": 0.07562015578150749, "chaser_new_fields": 92.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.4350004196167, "chaser_subspace_norm": 68.97872161865234, "collisions_per_episode": 0.550000011920929, "entropy": 0.002834049693774432, "explorer_grad_norm": 0.21107512712478638, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 120.04378509521484, "final_distance": 3.180603504180908, "max_param_abs": 2.619825839996338, "policy_loss": -0.00039526084764318625, "value_loss": 0.06872337684035301}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:57.568775+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014957.5687876, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5361426051116562, "elapsed_seconds": 12730.113378401962, "env_steps_per_second": 4084.8025822160957, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9040767873678148, "eta_seconds": 6854.676434524134, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 4.7623881740954666e-05, "chaser_grad_norm": 0.06886796839535236, "chaser_new_fields": 93.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 67.8337173461914, "collisions_per_episode": 0.45000001788139343, "entropy": 0.003157528582960367, "explorer_grad_norm": 0.4245411716401577, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 119.1957778930664, "final_distance": 3.7806153297424316, "max_param_abs": 2.649846315383911, "policy_loss": -0.00031039492675333236, "value_loss": 0.09159622341394424}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:04:02.105827+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015442.1058404, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6700703145152915, "elapsed_seconds": 13212.25313225505, "env_steps_per_second": 4087.115154353946, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7670708921740292, "eta_seconds": 6361.455211826506, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 4.650597816358015e-05, "chaser_grad_norm": 0.06667145155370235, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.630000114440918, "chaser_subspace_norm": 71.45598602294922, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0034325195010751486, "explorer_grad_norm": 0.2968932166695595, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 119.08173370361328, "final_distance": 3.7372214794158936, "max_param_abs": 2.6703202724456787, "policy_loss": -0.00023720909302937798, "value_loss": 0.04756408557295799}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:12:04.245579+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015924.245592, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.8039523158305544, "elapsed_seconds": 13694.228336989996, "env_steps_per_second": 4089.31402499959, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6302652782130949, "eta_seconds": 5868.955001567141, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -1.3820697354560707e-05, "chaser_grad_norm": 0.04255277942866087, "chaser_new_fields": 93.375, "chaser_partner_vision": 0.0, "chaser_return": 8.630000114440918, "chaser_subspace_norm": 72.71815490722656, "collisions_per_episode": 0.45000001788139343, "entropy": 0.0042581233428791165, "explorer_grad_norm": 0.27306586131453514, "explorer_new_fields": 95.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 121.31263732910156, "final_distance": 3.9743010997772217, "max_param_abs": 2.698016405105591, "policy_loss": -0.00027522624793263617, "value_loss": 0.04279269836843014}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:06.220785+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016406.2208, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9384114492808457, "elapsed_seconds": 14178.281217411044, "env_steps_per_second": 4090.763831710118, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4938802048996311, "eta_seconds": 5377.968737638672, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00015370788120971923, "chaser_grad_norm": 0.03278797212988138, "chaser_new_fields": 92.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.5600004196167, "chaser_subspace_norm": 78.8346939086914, "collisions_per_episode": 0.5, "entropy": 0.004240104230120778, "explorer_grad_norm": 0.3519405126571655, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 130.72972106933594, "final_distance": 4.106181621551514, "max_param_abs": 2.7262022495269775, "policy_loss": -0.00016195154929476985, "value_loss": 0.043409342877566814}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:10.273665+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016890.2736769, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.075734397235566, "elapsed_seconds": 14672.643830048037, "env_steps_per_second": 4089.242586065252, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3585781324118553, "eta_seconds": 4890.8812766826795, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 1.1531024195349238e-05, "chaser_grad_norm": 0.033809758722782135, "chaser_new_fields": 93.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 78.26759338378906, "collisions_per_episode": 0.550000011920929, "entropy": 0.0020056855864822865, "explorer_grad_norm": 0.1279930453747511, "explorer_new_fields": 95.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 126.9104232788086, "final_distance": 4.412608623504639, "max_param_abs": 2.740053176879883, "policy_loss": -9.81771981753976e-05, "value_loss": 0.0635758638381958}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:24.636278+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017384.6362925, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.212531428042211, "elapsed_seconds": 15165.113140951958, "env_steps_per_second": 4088.330856732934, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2229929952380612, "eta_seconds": 4402.77478285702, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0032838389529202816, "chaser_grad_norm": 0.06836945004761219, "chaser_new_fields": 93.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.690001487731934, "chaser_subspace_norm": 80.9717025756836, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0026452502934262156, "explorer_grad_norm": 0.13407007232308388, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 127.62902069091797, "final_distance": 4.132334232330322, "max_param_abs": 2.758204698562622, "policy_loss": -0.00015382529371521514, "value_loss": 0.04261405952274799}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:37.105588+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017877.1055987, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.3478986613297215, "elapsed_seconds": 15652.435180786997, "env_steps_per_second": 4088.8206378620575, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0869746653324304, "eta_seconds": 3913.1087951967493, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0012575061750773608, "chaser_grad_norm": 0.028763120993971825, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 80.18560028076172, "collisions_per_episode": 0.45000001788139343, "entropy": 0.0024889467167668045, "explorer_grad_norm": 0.20823930203914642, "explorer_new_fields": 95.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.4625015258789, "explorer_subspace_norm": 136.8090362548828, "final_distance": 3.9617085456848145, "max_param_abs": 2.774733066558838, "policy_loss": -0.0006444373294982597, "value_loss": 0.030278190039098263}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:44.427629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018364.4276435, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.481625659188624, "elapsed_seconds": 16133.852373079048, "env_steps_per_second": 4090.777482886085, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9506478671006173, "eta_seconds": 3422.332321562222, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": -9.181823818860124e-05, "chaser_grad_norm": 0.06706667691469193, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.644999504089355, "chaser_subspace_norm": 82.16859436035156, "collisions_per_episode": 0.5, "entropy": 0.0034693812485784292, "explorer_grad_norm": 0.3202550671994686, "explorer_new_fields": 95.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.3499984741211, "explorer_subspace_norm": 135.02574157714844, "final_distance": 4.19083309173584, "max_param_abs": 2.788562297821045, "policy_loss": -0.0003576974979360159, "value_loss": 0.04477046988904476}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:45.844820+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018845.844832, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.618225405726407, "elapsed_seconds": 16625.611460615066, "env_steps_per_second": 4090.075132640224, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.814980953951719, "eta_seconds": 2933.9314342261882, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0007968248343219742, "chaser_grad_norm": 0.10342094302177429, "chaser_new_fields": 93.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 84.98808288574219, "collisions_per_episode": 0.125, "entropy": 0.002209584228694439, "explorer_grad_norm": 0.29085009172558784, "explorer_new_fields": 95.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.4375, "explorer_subspace_norm": 134.9689483642578, "final_distance": 4.616744518280029, "max_param_abs": 2.800830602645874, "policy_loss": -0.0004560549503667133, "value_loss": 0.01913414103910327}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:08:57.603908+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019337.603923, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.7545037821358225, "elapsed_seconds": 17116.21361568896, "env_steps_per_second": 4089.6895523573635, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6792148260194032, "eta_seconds": 2445.1733736698516, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 8.018034570356214e-05, "chaser_grad_norm": 0.047528491355478764, "chaser_new_fields": 93.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.769999504089355, "chaser_subspace_norm": 83.74442291259766, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0029699402512051165, "explorer_grad_norm": 0.2107248269021511, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 131.8700714111328, "final_distance": 4.045284271240234, "max_param_abs": 2.8165388107299805, "policy_loss": -0.0002612216699064973, "value_loss": 0.08247810043394566}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:17:08.206063+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019828.2060747, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.888731290616658, "elapsed_seconds": 17599.432646219968, "env_steps_per_second": 4091.0409697476393, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5431923656240731, "eta_seconds": 1955.4925162466632, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0002844389394993929, "chaser_grad_norm": 0.036604116670787334, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.770000457763672, "chaser_subspace_norm": 85.01388549804688, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0026826930115930736, "explorer_grad_norm": 0.2773788385093212, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.3499984741211, "explorer_subspace_norm": 128.55809020996094, "final_distance": 3.861576795578003, "max_param_abs": 2.830378532409668, "policy_loss": -0.00023015607497656188, "value_loss": 0.06609669700264931}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:11.425095+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020311.42511, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.0254162651230585, "elapsed_seconds": 18091.49855444301, "env_steps_per_second": 4090.3189847602025, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40746618365862636, "eta_seconds": 1466.8782611710549, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0001132674199098016, "chaser_grad_norm": 0.03674895199947059, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.850000381469727, "chaser_subspace_norm": 86.63163757324219, "collisions_per_episode": 0.2750000059604645, "entropy": 0.0023306639050133526, "explorer_grad_norm": 0.16000011935830116, "explorer_new_fields": 95.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.2125015258789, "explorer_subspace_norm": 133.56771850585938, "final_distance": 3.9755799770355225, "max_param_abs": 2.8445000648498535, "policy_loss": -0.00010502004870716064, "value_loss": 0.024576107505708933}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:23.491002+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020803.4910142, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.163160475746392, "elapsed_seconds": 18587.37771268701, "env_steps_per_second": 4088.7962344535235, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2717452881971785, "eta_seconds": 978.2830375098428, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 1.165527411541023e-05, "chaser_grad_norm": 0.02973249088972807, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 84.8570327758789, "collisions_per_episode": 0.17499999701976776, "entropy": 0.001998357241973281, "explorer_grad_norm": 0.322164174169302, "explorer_new_fields": 95.75, "explorer_partner_vision": 0.0, "explorer_return": 93.4749984741211, "explorer_subspace_norm": 131.4634552001953, "final_distance": 4.490052223205566, "max_param_abs": 2.842074155807495, "policy_loss": -3.848600599987151e-05, "value_loss": 0.019126664847135544}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:39.370162+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021299.370177, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.2997301635391585, "elapsed_seconds": 19079.02858874097, "env_steps_per_second": 4088.258458086793, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13589051701382457, "eta_seconds": 489.2058612497685, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0015112189614507887, "chaser_grad_norm": 0.03277226723730564, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.86500072479248, "chaser_subspace_norm": 89.1612548828125, "collisions_per_episode": 0.20000000298023224, "entropy": 0.0019229634199291468, "explorer_grad_norm": 0.13945288211107254, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 139.1415252685547, "final_distance": 4.151286602020264, "max_param_abs": 2.8724098205566406, "policy_loss": -0.0001659054823957007, "value_loss": 0.02723497338593006}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:51.021035+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021791.021046, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.433286101221391, "elapsed_seconds": 19559.82996439701, "env_steps_per_second": 4090.0151047129125, "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-transformer/non_social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:51.822411+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022271.8224213, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.433294409401101, "elapsed_seconds": 19559.85987384396, "env_steps_per_second": 4090.0088505735375, "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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:51.852322+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022271.8523347, "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..6bc4d1baaab5d43d1ad1bc0c40392870dfdda02a --- /dev/null +++ b/non_social/seed_0004/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.433294409401101, + "elapsed_seconds": 19559.85987384396, + "env_steps_per_second": 4090.0088505735375, + "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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0017689612679874234, + "chaser_grad_norm": 0.025046652648597956, + "chaser_new_fields": 94.0, + "chaser_partner_vision": 0.0, + "chaser_return": 8.77500057220459, + "chaser_subspace_norm": 89.04515075683594, + "collisions_per_episode": 0.4000000059604645, + "entropy": 0.0025691293994896114, + "explorer_grad_norm": 0.1489477176219225, + "explorer_new_fields": 95.05000305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 92.38750457763672, + "explorer_subspace_norm": 140.36648559570312, + "final_distance": 3.9162888526916504, + "max_param_abs": 2.88704776763916, + "policy_loss": -0.0007027814439490854, + "value_loss": 0.05476926174014807 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:57:51.852322+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022271.8523347, + "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..5a9af1a4176de17394a88ad2ac6aac8bae4a97e6 --- /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:ac8c8ce4aab7af2d19fc15778486415942ca2cc3a405b65f8c5b23c7cfb69376 +size 29760360 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..b046bbd642d06bf04fe26b2862a6566e120912e5 --- /dev/null +++ b/non_social/seed_0005/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..70e479e5c52426afe3aaab8590345cee2e76d4a5 --- /dev/null +++ b/non_social/seed_0005/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006920439502573572, "elapsed_seconds": 2.491358220926486, "env_steps_per_second": 1605.549923090739, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.840186961196887, "eta_seconds": 49824.67306030879, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.008184439225479068, "chaser_grad_norm": 6.440325319766998, "chaser_new_fields": 34.04999923706055, "chaser_partner_vision": 0.0, "chaser_return": -3.2350003719329834, "chaser_subspace_norm": 0.033832646906375885, "collisions_per_episode": 1.225000023841858, "entropy": 2.7503525018692017, "explorer_grad_norm": 8.511807143688202, "explorer_new_fields": 34.72500228881836, "explorer_partner_vision": 0.0, "explorer_return": 1.8625000715255737, "explorer_subspace_norm": 0.039189036935567856, "final_distance": 5.033384799957275, "max_param_abs": 1.0011991262435913, "policy_loss": -0.008689469220115598, "value_loss": 7.674896478652954}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:48.329611+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022268.3296232, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13179484498415453, "elapsed_seconds": 474.4614419429563, "env_steps_per_second": 4215.305656471989, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.139998954382026, "eta_seconds": 18503.996235775296, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0009361219235302087, "chaser_grad_norm": 0.45476044714450836, "chaser_new_fields": 91.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.145001411437988, "chaser_subspace_norm": 11.66552734375, "collisions_per_episode": 1.2000000476837158, "entropy": 0.07734057679772377, "explorer_grad_norm": 10.573444128036499, "explorer_new_fields": 92.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 88.67500305175781, "explorer_subspace_norm": 15.748729705810547, "final_distance": 2.8839144706726074, "max_param_abs": 1.2961766719818115, "policy_loss": -0.0011633181998043085, "value_loss": 0.42136064916849136}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:05:40.299696+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022740.299709, "update": 500} +{"device": "cuda", "elapsed_hours": 0.26813214382108225, "elapsed_seconds": 965.275717755896, "env_steps_per_second": 4143.893735666871, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.094510732600561, "eta_seconds": 18340.23863736202, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00010687408780452046, "chaser_grad_norm": 0.22473064064979553, "chaser_new_fields": 90.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.930000305175781, "chaser_subspace_norm": 15.033252716064453, "collisions_per_episode": 2.5, "entropy": 0.021731394343078136, "explorer_grad_norm": 5.488495588302612, "explorer_new_fields": 94.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 90.9625015258789, "explorer_subspace_norm": 30.144948959350586, "final_distance": 3.47773814201355, "max_param_abs": 1.433841586112976, "policy_loss": -0.0012641563932476174, "value_loss": 0.3978065028786659}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:13:51.113971+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023231.1139865, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4040637830274904, "elapsed_seconds": 1454.6296188989654, "env_steps_per_second": 4124.761328963936, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.983453324005715, "eta_seconds": 17940.431966420572, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0015242458913891643, "chaser_grad_norm": 0.30585379526019096, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.315000534057617, "chaser_subspace_norm": 17.891708374023438, "collisions_per_episode": 1.9500000476837158, "entropy": 0.017732564825564623, "explorer_grad_norm": 4.706849932670593, "explorer_new_fields": 95.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.20000457763672, "explorer_subspace_norm": 39.83221435546875, "final_distance": 3.36601185798645, "max_param_abs": 1.5062170028686523, "policy_loss": -0.0018727443481338923, "value_loss": 0.41987208276987076}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:00.467872+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023720.4678838, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5393184688335815, "elapsed_seconds": 1941.5464878008934, "env_steps_per_second": 4120.426706373257, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.853866219502233, "eta_seconds": 17473.91839020804, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0005957404282430634, "chaser_grad_norm": 0.07774509489536285, "chaser_new_fields": 93.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.535000801086426, "chaser_subspace_norm": 20.057270050048828, "collisions_per_episode": 1.274999976158142, "entropy": 0.009850812144577503, "explorer_grad_norm": 2.4770524203777313, "explorer_new_fields": 96.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.26250457763672, "explorer_subspace_norm": 50.30194091796875, "final_distance": 3.6079468727111816, "max_param_abs": 1.5557539463043213, "policy_loss": -0.0008870726244012417, "value_loss": 0.19083719700574875}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:07.384743+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024207.3847592, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6764874987074937, "elapsed_seconds": 2435.3549953469774, "env_steps_per_second": 4106.177546643564, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.735412490952456, "eta_seconds": 17047.48496742884, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0020468066642520544, "chaser_grad_norm": 0.18242399487644434, "chaser_new_fields": 92.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.444999694824219, "chaser_subspace_norm": 23.863012313842773, "collisions_per_episode": 1.4250000715255737, "entropy": 0.006107415771111846, "explorer_grad_norm": 2.5438318848609924, "explorer_new_fields": 96.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.11250305175781, "explorer_subspace_norm": 55.08370590209961, "final_distance": 3.9846444129943848, "max_param_abs": 1.595381498336792, "policy_loss": -0.0010705109410906744, "value_loss": 0.18556535616517067}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:21.193247+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024701.19326, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8146118185836045, "elapsed_seconds": 2932.6025469009764, "env_steps_per_second": 4091.9285201743355, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.6161336386404255, "eta_seconds": 16618.081099105533, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0012350974433408268, "chaser_grad_norm": 0.10789668932557106, "chaser_new_fields": 92.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.445000648498535, "chaser_subspace_norm": 25.165456771850586, "collisions_per_episode": 1.0750000476837158, "entropy": 0.009079070296138525, "explorer_grad_norm": 1.4245552122592926, "explorer_new_fields": 97.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 61.17123031616211, "final_distance": 3.75791597366333, "max_param_abs": 1.6309267282485962, "policy_loss": -0.000544090077980286, "value_loss": 0.15173404291272163}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:46:38.440800+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025198.4408138, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9537632968552578, "elapsed_seconds": 3433.5478686789284, "env_steps_per_second": 4077.415121457607, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.496312685174788, "eta_seconds": 16186.725666629234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.002441858525618623, "chaser_grad_norm": 0.08779280353337526, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.555000305175781, "chaser_subspace_norm": 27.106536865234375, "collisions_per_episode": 1.350000023841858, "entropy": 0.005455998471006751, "explorer_grad_norm": 1.6851049661636353, "explorer_new_fields": 96.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.48750305175781, "explorer_subspace_norm": 64.252197265625, "final_distance": 4.024997234344482, "max_param_abs": 1.661129117012024, "policy_loss": -0.0019822102063358216, "value_loss": 0.2504864037036896}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:54:59.386121+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025699.3861337, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0913558711752476, "elapsed_seconds": 3928.8811362308916, "env_steps_per_second": 4072.4062258980275, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.36542348470099, "eta_seconds": 15715.524544923566, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0006985393587002342, "chaser_grad_norm": 0.044110307935625315, "chaser_new_fields": 93.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.595000267028809, "chaser_subspace_norm": 28.84015655517578, "collisions_per_episode": 1.524999976158142, "entropy": 0.004622174659743905, "explorer_grad_norm": 2.166859894990921, "explorer_new_fields": 96.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.26250457763672, "explorer_subspace_norm": 72.42479705810547, "final_distance": 3.289487600326538, "max_param_abs": 1.6978211402893066, "policy_loss": -0.0011017804748902904, "value_loss": 0.20734801143407822}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:03:14.719390+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026194.7194054, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.226758298336095, "elapsed_seconds": 4416.329874009942, "env_steps_per_second": 4075.7824966676117, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.225500805379883, "eta_seconds": 15211.802899367578, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0034291903907175403, "chaser_grad_norm": 0.09173675160855055, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.520000457763672, "chaser_subspace_norm": 26.406984329223633, "collisions_per_episode": 1.2000000476837158, "entropy": 0.006392874172888696, "explorer_grad_norm": 0.6193827465176582, "explorer_new_fields": 96.875, "explorer_partner_vision": 0.0, "explorer_return": 94.5625, "explorer_subspace_norm": 76.7430191040039, "final_distance": 3.7917816638946533, "max_param_abs": 1.7314627170562744, "policy_loss": -0.0021327687255165984, "value_loss": 0.21480071172118187}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:11:22.168129+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026682.1681428, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3627622604108183, "elapsed_seconds": 4905.944137478946, "env_steps_per_second": 4076.68726743341, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.088286781232455, "eta_seconds": 14717.832412436837, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0017705824037278495, "chaser_grad_norm": 0.11722816992551088, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.605000495910645, "chaser_subspace_norm": 30.065780639648438, "collisions_per_episode": 1.5500000715255737, "entropy": 0.00480334204621613, "explorer_grad_norm": 1.15304896235466, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 77.98207092285156, "final_distance": 3.5316147804260254, "max_param_abs": 1.7599422931671143, "policy_loss": -0.00027412737708054635, "value_loss": 0.16086706891655922}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:31.782392+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027171.7824078, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4970714932180513, "elapsed_seconds": 5389.457375584985, "env_steps_per_second": 4082.0436023231496, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.946824845756681, "eta_seconds": 14208.569444724051, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0004030766968986521, "chaser_grad_norm": 0.12016704119741917, "chaser_new_fields": 94.375, "chaser_partner_vision": 0.0, "chaser_return": 8.710000991821289, "chaser_subspace_norm": 32.22527313232422, "collisions_per_episode": 1.5500000715255737, "entropy": 0.007395019638352096, "explorer_grad_norm": 2.0366511791944504, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.9749984741211, "explorer_subspace_norm": 82.64091491699219, "final_distance": 3.48870587348938, "max_param_abs": 1.7974824905395508, "policy_loss": -0.00045716741646373293, "value_loss": 0.17801041528582573}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:35.295629+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027655.295642, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.631091633502478, "elapsed_seconds": 5871.929880608921, "env_steps_per_second": 4087.2422675304824, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.8058804781724485, "eta_seconds": 13701.169721420814, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": -1.0344755381197501e-06, "chaser_grad_norm": 0.13730387203395367, "chaser_new_fields": 94.25, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 33.81676483154297, "collisions_per_episode": 1.3250000476837158, "entropy": 0.0025091858115047216, "explorer_grad_norm": 0.5365075767040253, "explorer_new_fields": 97.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 87.90255737304688, "final_distance": 3.928429365158081, "max_param_abs": 1.8272534608840942, "policy_loss": -0.00017002964968426681, "value_loss": 0.12052592821419239}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:37.768134+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028137.7681503, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7660568435352697, "elapsed_seconds": 6357.804636726971, "env_steps_per_second": 4089.4619268114106, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6679642134963295, "eta_seconds": 13204.671168586787, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.000521471640998819, "chaser_grad_norm": 0.02714073657989502, "chaser_new_fields": 94.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.78499984741211, "chaser_subspace_norm": 34.732913970947266, "collisions_per_episode": 1.350000023841858, "entropy": 0.0043078402522951365, "explorer_grad_norm": 0.7792868465185165, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 91.30493927001953, "final_distance": 3.708338975906372, "max_param_abs": 1.8683478832244873, "policy_loss": -0.000430340119405237, "value_loss": 0.09126785956323147}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:43.642889+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028623.6429017, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.9039363565347027, "elapsed_seconds": 6854.17088352493, "env_steps_per_second": 4085.103869718564, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.535881804993019, "eta_seconds": 12729.17449797487, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00018198393097622334, "chaser_grad_norm": 0.036112076602876186, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 36.714996337890625, "collisions_per_episode": 1.2000000476837158, "entropy": 0.003180745872668922, "explorer_grad_norm": 1.101660594344139, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 92.3774185180664, "final_distance": 3.520420789718628, "max_param_abs": 1.898836612701416, "policy_loss": -0.00017902375321154906, "value_loss": 0.09390992298722267}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:52:00.009138+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029120.009152, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.040748056849148, "elapsed_seconds": 7346.693004656932, "env_steps_per_second": 4083.4699341572536, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.4012467614152464, "eta_seconds": 12244.488341094888, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00042230356906974853, "chaser_grad_norm": 0.07996394485235214, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.744999885559082, "chaser_subspace_norm": 37.123470306396484, "collisions_per_episode": 1.3250000476837158, "entropy": 0.0034858802682720125, "explorer_grad_norm": 0.5679749622941017, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 100.16153717041016, "final_distance": 3.3677127361297607, "max_param_abs": 1.9363983869552612, "policy_loss": -0.0003527298206460472, "value_loss": 0.09695837274193764}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:00:12.531257+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029612.5312696, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.176398255478028, "elapsed_seconds": 7835.033719720901, "env_steps_per_second": 4084.219818921201, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2645973832170423, "eta_seconds": 11752.550579581351, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0019244914414586756, "chaser_grad_norm": 0.12822234071791172, "chaser_new_fields": 94.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.870000839233398, "chaser_subspace_norm": 41.26630783081055, "collisions_per_episode": 0.875, "entropy": 0.0027616055449470878, "explorer_grad_norm": 1.0141589939594269, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 100.3052749633789, "final_distance": 3.0593831539154053, "max_param_abs": 1.9546184539794922, "policy_loss": -0.0003915233662130646, "value_loss": 0.10390745662152767}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:08:20.871973+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030100.8719869, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.310681776522219, "elapsed_seconds": 8318.454395479988, "env_steps_per_second": 4087.297757919384, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1262165211771196, "eta_seconds": 11254.379476237631, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": -1.4335217940003586e-05, "chaser_grad_norm": 0.04571158718317747, "chaser_new_fields": 94.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.805000305175781, "chaser_subspace_norm": 41.69701385498047, "collisions_per_episode": 1.375, "entropy": 0.0025362682645209134, "explorer_grad_norm": 0.7047380357980728, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 105.92369079589844, "final_distance": 3.0623576641082764, "max_param_abs": 1.9880163669586182, "policy_loss": -8.808446089059174e-05, "value_loss": 0.1007557287812233}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:16:24.292648+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030584.2926602, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4455032916399717, "elapsed_seconds": 8803.811849903897, "env_steps_per_second": 4089.1378204990806, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.988948467559965, "eta_seconds": 10760.214483215874, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00029666933588012867, "chaser_grad_norm": 0.040595004335045815, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.690000534057617, "chaser_subspace_norm": 42.999568939208984, "collisions_per_episode": 1.3000000715255737, "entropy": 0.0029416828765533864, "explorer_grad_norm": 0.41351157054305077, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 105.0737075805664, "final_distance": 2.9586522579193115, "max_param_abs": 2.0143489837646484, "policy_loss": -0.00024724579969159777, "value_loss": 0.11778833530843258}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:29.650106+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031069.6501222, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5814078495824813, "elapsed_seconds": 9293.068258496933, "env_steps_per_second": 4089.069287235188, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.853134991643795, "eta_seconds": 10271.285969917662, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0017763836128132163, "chaser_grad_norm": 0.06757924798876047, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.824999809265137, "chaser_subspace_norm": 42.66282272338867, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0017758093017619103, "explorer_grad_norm": 1.1538244783878326, "explorer_new_fields": 97.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 109.50704956054688, "final_distance": 3.04056715965271, "max_param_abs": 2.048684597015381, "policy_loss": -0.0013928900831245983, "value_loss": 0.11639217287302017}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:38.906511+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031558.9065232, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.718024756178331, "elapsed_seconds": 9784.889122241992, "env_steps_per_second": 4087.935949021248, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.718024756178331, "eta_seconds": 9784.889122241992, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": -1.4062203634868098e-05, "chaser_grad_norm": 0.05896423850208521, "chaser_new_fields": 94.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 46.2486572265625, "collisions_per_episode": 1.0750000476837158, "entropy": 0.003515815536957234, "explorer_grad_norm": 0.31926171854138374, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 108.01021575927734, "final_distance": 2.9094836711883545, "max_param_abs": 2.0774147510528564, "policy_loss": -0.000163217312160846, "value_loss": 0.08966618403792381}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:40:50.727376+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032050.7273905, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.853866963133882, "elapsed_seconds": 10273.921067281975, "env_steps_per_second": 4088.0205059927857, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5820701095020837, "eta_seconds": 9295.452394207501, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0030544491095406134, "chaser_grad_norm": 0.15306414104998112, "chaser_new_fields": 94.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.885001182556152, "chaser_subspace_norm": 48.492122650146484, "collisions_per_episode": 1.024999976158142, "entropy": 0.00357656559208408, "explorer_grad_norm": 0.9481620192527771, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 113.7264633178711, "final_distance": 3.156186103820801, "max_param_abs": 2.099663019180298, "policy_loss": -0.0007126822820868739, "value_loss": 0.11961465328931808}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:48:59.759321+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032539.7593343, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9875429796324977, "elapsed_seconds": 10755.154726676992, "env_steps_per_second": 4091.0615531046506, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4443533469720435, "eta_seconds": 8799.672049099358, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00020237287433322315, "chaser_grad_norm": 0.03940554801374674, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.795001029968262, "chaser_subspace_norm": 48.79186248779297, "collisions_per_episode": 1.3000000715255737, "entropy": 0.0031767572509124875, "explorer_grad_norm": 0.48446035385131836, "explorer_new_fields": 97.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.7125015258789, "explorer_subspace_norm": 116.57722473144531, "final_distance": 3.3267982006073, "max_param_abs": 2.1329665184020996, "policy_loss": -0.000321724902139664, "value_loss": 0.10522451065480709}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:00.992980+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033020.992996, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.121035432090252, "elapsed_seconds": 11235.727555524907, "env_steps_per_second": 4094.0828951820367, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.306852275892795, "eta_seconds": 8304.668193214062, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004090758606384304, "chaser_grad_norm": 0.19424590468406677, "chaser_new_fields": 94.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 50.733299255371094, "collisions_per_episode": 1.25, "entropy": 0.002097469550790265, "explorer_grad_norm": 0.6892369091510773, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 116.6259765625, "final_distance": 2.8124120235443115, "max_param_abs": 2.1573104858398438, "policy_loss": -0.0005452695031418209, "value_loss": 0.10252008028328419}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:01.565808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033501.5658207, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.254765800418035, "elapsed_seconds": 11717.156881504925, "env_steps_per_second": 4096.556910982913, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1698438669453566, "eta_seconds": 7811.437921003284, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.001365038953287545, "chaser_grad_norm": 0.10101035609841347, "chaser_new_fields": 94.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 50.75255584716797, "collisions_per_episode": 0.8500000238418579, "entropy": 0.0037602168740704656, "explorer_grad_norm": 0.43663299828767776, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 119.86353302001953, "final_distance": 3.1305291652679443, "max_param_abs": 2.1793112754821777, "policy_loss": -6.250381696659701e-05, "value_loss": 0.10334442555904388}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:13:02.995136+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033982.995153, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3898560839169334, "elapsed_seconds": 12203.48190210096, "env_steps_per_second": 4097.191309915571, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0339136503501605, "eta_seconds": 7322.089141260577, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.000989114510704936, "chaser_grad_norm": 0.08919528219848871, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 51.018123626708984, "collisions_per_episode": 1.274999976158142, "entropy": 0.0036947763292118907, "explorer_grad_norm": 0.49894385039806366, "explorer_new_fields": 97.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 125.3757095336914, "final_distance": 2.5259768962860107, "max_param_abs": 2.2026240825653076, "policy_loss": -0.0006352567888097127, "value_loss": 0.11523559875786304}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:21:09.320156+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034469.3201694, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5251186574160887, "elapsed_seconds": 12690.427166697918, "env_steps_per_second": 4097.576804700305, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8981408155317399, "eta_seconds": 6833.306935914264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 9.058276429174672e-05, "chaser_grad_norm": 0.048134016804397106, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 52.3425178527832, "collisions_per_episode": 1.5, "entropy": 0.0022945620585232973, "explorer_grad_norm": 1.0205857157707214, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 129.28347778320312, "final_distance": 2.797969102859497, "max_param_abs": 2.2236173152923584, "policy_loss": -8.505058767127593e-05, "value_loss": 0.11947360821068287}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:29:16.265420+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034956.265435, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6616521055691473, "elapsed_seconds": 13181.94758004893, "env_steps_per_second": 4096.511511070624, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7630176804592192, "eta_seconds": 6346.863649653189, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.00023019519146316725, "chaser_grad_norm": 0.04138797149062157, "chaser_new_fields": 94.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 56.570579528808594, "collisions_per_episode": 1.2000000476837158, "entropy": 0.002282455679960549, "explorer_grad_norm": 0.938120424747467, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.3125, "explorer_subspace_norm": 131.00253295898438, "final_distance": 2.6901204586029053, "max_param_abs": 2.2481918334960938, "policy_loss": -0.00024209464679003645, "value_loss": 0.11030559986829758}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:37:27.785834+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035447.7858477, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.797376491899712, "elapsed_seconds": 13670.555370838963, "env_steps_per_second": 4096.395390011377, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6274470679570197, "eta_seconds": 5858.809444645271, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0015203803531047555, "chaser_grad_norm": 0.07362508960068226, "chaser_new_fields": 94.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.690000534057617, "chaser_subspace_norm": 58.064247131347656, "collisions_per_episode": 1.475000023841858, "entropy": 0.004419806529767811, "explorer_grad_norm": 1.720163807272911, "explorer_new_fields": 97.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 126.17747497558594, "final_distance": 2.929892063140869, "max_param_abs": 2.2679409980773926, "policy_loss": -0.0007249031414975349, "value_loss": 0.13401664420962334}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:45:36.393624+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035936.3936381, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.93200999457166, "elapsed_seconds": 14155.235980457976, "env_steps_per_second": 4097.423743417062, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4914520669064917, "eta_seconds": 5369.22744086337, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0016421328080014064, "chaser_grad_norm": 0.1113754017278552, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 60.42548370361328, "collisions_per_episode": 1.024999976158142, "entropy": 0.003006738726980984, "explorer_grad_norm": 0.8497034683823586, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 130.97738647460938, "final_distance": 2.5862414836883545, "max_param_abs": 2.383998394012451, "policy_loss": -0.0005429268329173453, "value_loss": 0.11908523179590702}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:53:41.074235+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036421.0742488, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.069114904249695, "elapsed_seconds": 14648.813655298902, "env_steps_per_second": 4095.8948220558636, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3563716347498982, "eta_seconds": 4882.937885099634, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004713547696755249, "chaser_grad_norm": 0.031762602273374796, "chaser_new_fields": 94.625, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 60.8377799987793, "collisions_per_episode": 1.149999976158142, "entropy": 0.0015576633741147816, "explorer_grad_norm": 0.5975871719419956, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 135.2548065185547, "final_distance": 2.847369432449341, "max_param_abs": 2.522535562515259, "policy_loss": -0.00016278053342588805, "value_loss": 0.07293580658733845}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:01:54.651910+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036914.6519248, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.205038526879992, "elapsed_seconds": 15138.138696767972, "env_steps_per_second": 4095.6157980793996, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2208176368361268, "eta_seconds": 4394.943492610057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0004081863581145151, "chaser_grad_norm": 0.05001601576805115, "chaser_new_fields": 94.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.84000015258789, "chaser_subspace_norm": 59.43809509277344, "collisions_per_episode": 0.7750000357627869, "entropy": 0.0034388050553388894, "explorer_grad_norm": 0.8913935273885727, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 139.73411560058594, "final_distance": 3.0137009620666504, "max_param_abs": 2.671187400817871, "policy_loss": -0.00016906086285572997, "value_loss": 0.10254292003810406}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:10:03.976952+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037403.976966, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.3405951427749825, "elapsed_seconds": 15626.142513989937, "env_steps_per_second": 4095.700518710962, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0851487856937456, "eta_seconds": 3906.535628497484, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0013712222712034539, "chaser_grad_norm": 0.06847671559080482, "chaser_new_fields": 94.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.74000072479248, "chaser_subspace_norm": 59.01997756958008, "collisions_per_episode": 1.0500000715255737, "entropy": 0.003109973040409386, "explorer_grad_norm": 0.5139616504311562, "explorer_new_fields": 97.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 139.3917236328125, "final_distance": 2.5476038455963135, "max_param_abs": 2.8179855346679688, "policy_loss": -0.00032580759574152296, "value_loss": 0.09241721779108047}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:18:11.980771+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037891.9807906, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.476565459541388, "elapsed_seconds": 16115.635654348996, "env_steps_per_second": 4095.4015972797893, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9495744914178701, "eta_seconds": 3418.4681691043324, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0002272066812404605, "chaser_grad_norm": 0.03070598654448986, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 62.595947265625, "collisions_per_episode": 0.9750000238418579, "entropy": 0.002077111799735576, "explorer_grad_norm": 0.3588371053338051, "explorer_new_fields": 97.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 132.83013916015625, "final_distance": 3.009052038192749, "max_param_abs": 2.94905161857605, "policy_loss": -7.086706634318318e-05, "value_loss": 0.0682746060192585}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:26:21.473906+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038381.4739184, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.613364270519976, "elapsed_seconds": 16608.111373871914, "env_steps_per_second": 4094.3848743077697, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8141231065623488, "eta_seconds": 2930.8431836244554, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0020441413901413075, "chaser_grad_norm": 0.045831756200641394, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 65.60694885253906, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0020277429139241576, "explorer_grad_norm": 0.5332756787538528, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 145.29444885253906, "final_distance": 3.0503880977630615, "max_param_abs": 3.096259832382202, "policy_loss": -6.159306025033118e-05, "value_loss": 0.07014592923223972}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:33.949628+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038873.9496448, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.749466954409694, "elapsed_seconds": 17098.0810358749, "env_steps_per_second": 4094.0266836452115, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.678495279201385, "eta_seconds": 2442.583005124986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0002671154577380541, "chaser_grad_norm": 0.0547892265021801, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.795001029968262, "chaser_subspace_norm": 66.88868713378906, "collisions_per_episode": 0.949999988079071, "entropy": 0.0021956649725325406, "explorer_grad_norm": 0.6824228912591934, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 147.3888702392578, "final_distance": 2.673384666442871, "max_param_abs": 3.2412221431732178, "policy_loss": -0.000199659829901222, "value_loss": 0.08626480400562286}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:43.919290+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039363.9193044, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.8856042489344444, "elapsed_seconds": 17588.175296164, "env_steps_per_second": 4093.659449465646, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5428449165482716, "eta_seconds": 1954.2416995737779, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00024259486997357804, "chaser_grad_norm": 0.06348314369097352, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.75, "chaser_subspace_norm": 68.30133819580078, "collisions_per_episode": 1.225000023841858, "entropy": 0.0032776488224044442, "explorer_grad_norm": 0.3755941763520241, "explorer_new_fields": 97.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 144.0692901611328, "final_distance": 2.865546464920044, "max_param_abs": 3.3910720348358154, "policy_loss": 1.6978267027401017e-05, "value_loss": 0.09523412771522999}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:50:54.013551+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039854.0135665, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.021417094115556, "elapsed_seconds": 18077.101538816, "env_steps_per_second": 4093.576608014494, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40714192654990994, "eta_seconds": 1465.7109355796758, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 7.123219962856808e-05, "chaser_grad_norm": 0.036403757985681295, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.760000228881836, "chaser_subspace_norm": 70.68766784667969, "collisions_per_episode": 1.225000023841858, "entropy": 0.0030727192061021924, "explorer_grad_norm": 0.42910633981227875, "explorer_new_fields": 97.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 145.63185119628906, "final_distance": 2.8040151596069336, "max_param_abs": 3.5327699184417725, "policy_loss": -0.0001866714523055535, "value_loss": 0.10456709191203117}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:59:02.939791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040342.9398034, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.155799050785556, "elapsed_seconds": 18560.876582828, "env_steps_per_second": 4094.6341979512463, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27135784477818714, "eta_seconds": 976.8882412014738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0011629241046633432, "chaser_grad_norm": 0.09430215880274773, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 69.79297637939453, "collisions_per_episode": 1.225000023841858, "entropy": 0.0034293356584385037, "explorer_grad_norm": 0.4956182390451431, "explorer_new_fields": 97.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 145.0971221923828, "final_distance": 3.0578739643096924, "max_param_abs": 3.6705596446990967, "policy_loss": -0.0006123671865765346, "value_loss": 0.08534514717757702}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:07:06.714836+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040826.7148514, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.29110973319637, "elapsed_seconds": 19047.99503950693, "env_steps_per_second": 4094.919168039592, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13566948033836845, "eta_seconds": 488.41012921812643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0001795099746915163, "chaser_grad_norm": 0.05988466367125511, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 70.86493682861328, "collisions_per_episode": 0.8500000238418579, "entropy": 0.0021517517743632197, "explorer_grad_norm": 0.41555262356996536, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 155.8712921142578, "final_distance": 3.2743000984191895, "max_param_abs": 3.82509183883667, "policy_loss": -6.89809348913828e-05, "value_loss": 0.062252127565443516}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:15:13.833292+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041313.8333035, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.423251158835806, "elapsed_seconds": 19523.704171808902, "env_steps_per_second": 4097.583086488033, "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-transformer/non_social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:09.542424+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041789.5424345, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.423259738547204, "elapsed_seconds": 19523.735058769933, "env_steps_per_second": 4097.5766040250855, "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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:09.573312+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041789.573325, "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..7a2b2dcf293e839bc015291cf3313c2d620212d9 --- /dev/null +++ b/non_social/seed_0005/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.423259738547204, + "elapsed_seconds": 19523.735058769933, + "env_steps_per_second": 4097.5766040250855, + "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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 4.4411225102368446e-05, + "chaser_grad_norm": 0.09301618672907352, + "chaser_new_fields": 94.4000015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.815000534057617, + "chaser_subspace_norm": 69.54359436035156, + "collisions_per_episode": 0.925000011920929, + "entropy": 0.002813033468555659, + "explorer_grad_norm": 0.6812433376908302, + "explorer_new_fields": 97.95000457763672, + "explorer_partner_vision": 0.0, + "explorer_return": 96.0625, + "explorer_subspace_norm": 156.21910095214844, + "final_distance": 3.1009461879730225, + "max_param_abs": 3.969390392303467, + "policy_loss": -0.0003465476266564238, + "value_loss": 0.07470510713756084 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:09.573312+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041789.573325, + "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..dc24338fbb6c4d3506810e533277a423d12ed15f --- /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:c3eb7fb2c91b2c029361c2a8a5a023ab29aba5b6cae2aca1cd33fac54dfe1d6a +size 29760368 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..758f94ddd8b54a9e2612e02952cf442b14459d21 --- /dev/null +++ b/non_social/seed_0006/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..a92882fef39b14d3aea6ad6a5fac56de699eedc9 --- /dev/null +++ b/non_social/seed_0006/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006340497011034232, "elapsed_seconds": 2.2825789239723235, "env_steps_per_second": 1752.4038086879752, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 12.680359972367361, "eta_seconds": 45649.2959005225, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.009081356627168358, "chaser_grad_norm": 9.28392294794321, "chaser_new_fields": 34.45000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.135000228881836, "chaser_subspace_norm": 0.032425787299871445, "collisions_per_episode": 1.350000023841858, "entropy": 2.7488085627555847, "explorer_grad_norm": 7.50808984041214, "explorer_new_fields": 32.82500076293945, "explorer_partner_vision": 0.0, "explorer_return": -0.949999988079071, "explorer_subspace_norm": 0.0343438945710659, "final_distance": 5.18532657623291, "max_param_abs": 1.0012028217315674, "policy_loss": -0.008591297162709566, "value_loss": 7.798083543777466}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:10.457463+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022290.4574733, "update": 1} +{"device": "cuda", "elapsed_hours": 0.1324592853458469, "elapsed_seconds": 476.8534272450488, "env_steps_per_second": 4194.160900876205, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.165912128488029, "eta_seconds": 18597.283662556903, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0017517860296626386, "chaser_grad_norm": 0.31775665283203125, "chaser_new_fields": 91.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.265000343322754, "chaser_subspace_norm": 11.57347297668457, "collisions_per_episode": 0.75, "entropy": 0.0558307059109211, "explorer_grad_norm": 10.75618839263916, "explorer_new_fields": 93.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 90.4375, "explorer_subspace_norm": 15.667323112487793, "final_distance": 2.812631368637085, "max_param_abs": 1.32639741897583, "policy_loss": -0.0034992986684585503, "value_loss": 0.35839923471212387}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:06:05.028314+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022765.028328, "update": 500} +{"device": "cuda", "elapsed_hours": 0.26846497917001316, "elapsed_seconds": 966.4739250120474, "env_steps_per_second": 4138.756252477416, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.100834604230251, "eta_seconds": 18363.0045752289, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0014362353670867023, "chaser_grad_norm": 0.4267645925283432, "chaser_new_fields": 92.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.410000801086426, "chaser_subspace_norm": 13.562819480895996, "collisions_per_episode": 0.5, "entropy": 0.018285356927663088, "explorer_grad_norm": 4.223830282688141, "explorer_new_fields": 96.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.8125, "explorer_subspace_norm": 28.409225463867188, "final_distance": 2.7266013622283936, "max_param_abs": 1.5229872465133667, "policy_loss": -0.002394474568310656, "value_loss": 0.1739034280180931}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:14:14.648810+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023254.648824, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4027435065277839, "elapsed_seconds": 1449.876623500022, "env_steps_per_second": 4138.28314957994, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.967169913842668, "eta_seconds": 17881.811689833605, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004477103975858654, "chaser_grad_norm": 0.32416776195168495, "chaser_new_fields": 90.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.040000915527344, "chaser_subspace_norm": 16.41919708251953, "collisions_per_episode": 0.949999988079071, "entropy": 0.01877806894481182, "explorer_grad_norm": 3.8259685933589935, "explorer_new_fields": 96.625, "explorer_partner_vision": 0.0, "explorer_return": 94.4124984741211, "explorer_subspace_norm": 39.73762893676758, "final_distance": 2.208803176879883, "max_param_abs": 1.6338715553283691, "policy_loss": -0.0017088156484490469, "value_loss": 0.17939159274101257}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:18.051506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023738.0515182, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5402188905961035, "elapsed_seconds": 1944.7880061459728, "env_steps_per_second": 4113.558894192158, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.861970015364933, "eta_seconds": 17503.09205531376, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0007497676800674591, "chaser_grad_norm": 0.1537339510396123, "chaser_new_fields": 92.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.440001487731934, "chaser_subspace_norm": 20.747173309326172, "collisions_per_episode": 0.699999988079071, "entropy": 0.008710291935130954, "explorer_grad_norm": 2.2933662235736847, "explorer_new_fields": 96.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 46.05669021606445, "final_distance": 2.3678863048553467, "max_param_abs": 1.6998072862625122, "policy_loss": -0.0007282829068131491, "value_loss": 0.10590462386608124}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:32.962892+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024232.9629066, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6760303580455689, "elapsed_seconds": 2433.709288964048, "env_steps_per_second": 4108.954198164185, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.732212506318982, "eta_seconds": 17035.965022748336, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002755190706777988, "chaser_grad_norm": 0.10564791224896908, "chaser_new_fields": 93.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.520000457763672, "chaser_subspace_norm": 23.54489517211914, "collisions_per_episode": 1.0500000715255737, "entropy": 0.009279369376599789, "explorer_grad_norm": 2.4777910113334656, "explorer_new_fields": 96.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.5250015258789, "explorer_subspace_norm": 52.84782028198242, "final_distance": 2.5959105491638184, "max_param_abs": 1.7529109716415405, "policy_loss": -0.0013189362334102839, "value_loss": 0.14943654462695122}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:41.884172+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024721.8841834, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8094591433922242, "elapsed_seconds": 2914.0529162120074, "env_steps_per_second": 4117.976009714629, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.58693514588927, "eta_seconds": 16512.966525201373, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0026490730582038324, "chaser_grad_norm": 0.1542376298457384, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 28.20026397705078, "collisions_per_episode": 0.6500000357627869, "entropy": 0.005760655738413334, "explorer_grad_norm": 2.126728057861328, "explorer_new_fields": 96.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.8125, "explorer_subspace_norm": 61.04405212402344, "final_distance": 3.0504865646362305, "max_param_abs": 1.7990180253982544, "policy_loss": -0.0018209304869998277, "value_loss": 0.1942107118666172}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:46:42.227800+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025202.227815, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9441227549747175, "elapsed_seconds": 3398.8419179089833, "env_steps_per_second": 4119.050058266023, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.450864416309383, "eta_seconds": 16023.111898713778, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.002258445791657515, "chaser_grad_norm": 0.18437314964830875, "chaser_new_fields": 93.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 29.267290115356445, "collisions_per_episode": 0.949999988079071, "entropy": 0.006399786798283458, "explorer_grad_norm": 0.8159359097480774, "explorer_new_fields": 96.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.0374984741211, "explorer_subspace_norm": 63.45840835571289, "final_distance": 2.536119222640991, "max_param_abs": 1.8377091884613037, "policy_loss": -0.001426826559240979, "value_loss": 0.11861684918403625}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:54:47.016801+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025687.016813, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0790977155599912, "elapsed_seconds": 3884.751776015968, "env_steps_per_second": 4118.66727207186, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.316390862239965, "eta_seconds": 15539.007104063872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0009261028763427048, "chaser_grad_norm": 0.07158619537949562, "chaser_new_fields": 93.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 31.17525863647461, "collisions_per_episode": 0.75, "entropy": 0.006486838683485985, "explorer_grad_norm": 1.674952894449234, "explorer_new_fields": 96.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.7125015258789, "explorer_subspace_norm": 68.22615051269531, "final_distance": 2.0703301429748535, "max_param_abs": 1.878548264503479, "policy_loss": -0.0013943439025752014, "value_loss": 0.15084588155150414}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:02:52.926661+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026172.9266772, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2143827589272083, "elapsed_seconds": 4371.777932137949, "env_steps_per_second": 4117.318006406008, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.182873947415939, "eta_seconds": 15058.34621069738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 6.878799733930907e-05, "chaser_grad_norm": 0.03787218267098069, "chaser_new_fields": 93.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.65000057220459, "chaser_subspace_norm": 34.393680572509766, "collisions_per_episode": 1.024999976158142, "entropy": 0.004154657945036888, "explorer_grad_norm": 1.4655900299549103, "explorer_new_fields": 96.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.86250305175781, "explorer_subspace_norm": 72.24091339111328, "final_distance": 2.4027812480926514, "max_param_abs": 1.914402723312378, "policy_loss": -0.00025126971033806456, "value_loss": 0.09570373967289925}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:10:59.952815+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026659.9528258, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.351526881251096, "elapsed_seconds": 4865.496772503946, "env_steps_per_second": 4110.577179502955, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.0545806437532885, "eta_seconds": 14596.490317511838, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 1.0621170141389413e-05, "chaser_grad_norm": 0.0821846779435873, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 33.195072174072266, "collisions_per_episode": 0.8500000238418579, "entropy": 0.004661013721488416, "explorer_grad_norm": 1.5245185494422913, "explorer_new_fields": 97.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 74.63639831542969, "final_distance": 2.5133373737335205, "max_param_abs": 1.9484796524047852, "policy_loss": -0.0004091162831763029, "value_loss": 0.06395275145769119}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:13.671657+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027153.6716738, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4883398381705693, "elapsed_seconds": 5358.02341741405, "env_steps_per_second": 4105.991759666084, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9238050279042285, "eta_seconds": 14125.698100455222, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00039155774840083507, "chaser_grad_norm": 0.04648858401924372, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.610000610351562, "chaser_subspace_norm": 36.63489532470703, "collisions_per_episode": 0.925000011920929, "entropy": 0.0043410195503383875, "explorer_grad_norm": 1.281730055809021, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 80.35256958007812, "final_distance": 2.188136339187622, "max_param_abs": 1.9779452085494995, "policy_loss": -0.0006414571033133143, "value_loss": 0.07625494711101055}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:26.198302+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027646.1983166, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6236991062505532, "elapsed_seconds": 5845.316782501992, "env_steps_per_second": 4105.851041614069, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7886312479179574, "eta_seconds": 13639.072492504647, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0023648358524350765, "chaser_grad_norm": 0.035658171866089106, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 40.32418441772461, "collisions_per_episode": 0.574999988079071, "entropy": 0.004197318048682064, "explorer_grad_norm": 1.142631858587265, "explorer_new_fields": 97.125, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 87.25827026367188, "final_distance": 2.5605618953704834, "max_param_abs": 2.0082149505615234, "policy_loss": -0.0031194363399400515, "value_loss": 0.11602420918643475}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:33.491668+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028133.4916828, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7596862807822051, "elapsed_seconds": 6334.870610815939, "env_steps_per_second": 4104.266937292847, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6547330447015036, "eta_seconds": 13157.038960925413, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003027150061973849, "chaser_grad_norm": 0.04697199258953333, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 39.59521484375, "collisions_per_episode": 1.2000000476837158, "entropy": 0.006377081153914332, "explorer_grad_norm": 0.44872088730335236, "explorer_new_fields": 96.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 94.75, "explorer_subspace_norm": 88.96295928955078, "final_distance": 2.341438055038452, "max_param_abs": 2.0350699424743652, "policy_loss": -0.00037196256092286717, "value_loss": 0.1247739065438509}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:43.045494+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028623.0455053, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8960810574647298, "elapsed_seconds": 6825.891806873027, "env_steps_per_second": 4102.028100094796, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5212933924344982, "eta_seconds": 12676.656212764194, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0015827478717496124, "chaser_grad_norm": 0.1372872330248356, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.675000190734863, "chaser_subspace_norm": 41.84708786010742, "collisions_per_episode": 0.6500000357627869, "entropy": 0.003968516713939607, "explorer_grad_norm": 0.7732554823160172, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.61250305175781, "explorer_subspace_norm": 88.53568267822266, "final_distance": 2.46395206451416, "max_param_abs": 2.065558433532715, "policy_loss": -0.001052412109914691, "value_loss": 0.0816511232405901}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:51:54.066690+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029114.0667057, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0298366709441567, "elapsed_seconds": 7307.412015398964, "env_steps_per_second": 4105.420624535851, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3830611182402617, "eta_seconds": 12179.020025664942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0019360221052711346, "chaser_grad_norm": 0.05764288455247879, "chaser_new_fields": 93.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 43.68484878540039, "collisions_per_episode": 0.699999988079071, "entropy": 0.0032946523278951645, "explorer_grad_norm": 0.9634853154420853, "explorer_new_fields": 97.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 96.0250015258789, "explorer_subspace_norm": 93.03872680664062, "final_distance": 2.0583181381225586, "max_param_abs": 2.0885467529296875, "policy_loss": -0.00018088604579613587, "value_loss": 0.06554706953465939}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:59:55.586898+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029595.5869105, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.164390133535554, "elapsed_seconds": 7791.804480727995, "env_steps_per_second": 4106.879231781008, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2465852003033313, "eta_seconds": 11687.706721091992, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0005512295202090844, "chaser_grad_norm": 0.0987094221636653, "chaser_new_fields": 93.875, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 48.18280792236328, "collisions_per_episode": 0.949999988079071, "entropy": 0.0032457555062137544, "explorer_grad_norm": 0.5241964012384415, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 93.68035125732422, "final_distance": 2.1907718181610107, "max_param_abs": 2.120203733444214, "policy_loss": -0.000570195695850817, "value_loss": 0.08561011590063572}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:07:59.979365+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030079.9793801, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.297568872478894, "elapsed_seconds": 8271.247940924019, "env_steps_per_second": 4110.625173231321, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.108475533353798, "eta_seconds": 11190.511920073672, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 9.063592311420177e-05, "chaser_grad_norm": 0.03654113505035639, "chaser_new_fields": 93.875, "chaser_partner_vision": 0.0, "chaser_return": 8.690001487731934, "chaser_subspace_norm": 48.02528381347656, "collisions_per_episode": 1.024999976158142, "entropy": 0.004057165002450347, "explorer_grad_norm": 0.4029681459069252, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 96.40164947509766, "final_distance": 2.4163246154785156, "max_param_abs": 2.141502857208252, "policy_loss": -0.00027873397825173285, "value_loss": 0.11679401993751526}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:15:59.422823+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030559.422835, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.432162019604455, "elapsed_seconds": 8755.783270576037, "env_steps_per_second": 4111.568192988357, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9726424684054447, "eta_seconds": 10701.512886259601, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": -2.4071496259894798e-05, "chaser_grad_norm": 0.025664323708042502, "chaser_new_fields": 93.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 51.08539962768555, "collisions_per_episode": 0.8500000238418579, "entropy": 0.002273507008794695, "explorer_grad_norm": 0.4124542959034443, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 96.16495513916016, "final_distance": 2.28855562210083, "max_param_abs": 2.1810531616210938, "policy_loss": -5.639028652515421e-05, "value_loss": 0.1025311816483736}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:03.958154+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031043.9581697, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.56710545852056, "elapsed_seconds": 9241.579650674015, "env_steps_per_second": 4111.851159258098, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8373270857332504, "eta_seconds": 10214.377508639702, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00022325854453769667, "chaser_grad_norm": 0.019994712667539716, "chaser_new_fields": 94.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 53.11989212036133, "collisions_per_episode": 1.024999976158142, "entropy": 0.001609528815606609, "explorer_grad_norm": 0.5562155470252037, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.1624984741211, "explorer_subspace_norm": 97.70769500732422, "final_distance": 2.372077703475952, "max_param_abs": 2.2171342372894287, "policy_loss": -0.00012136198038970036, "value_loss": 0.11478429846465588}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:09.754534+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031529.754543, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7005695298522348, "elapsed_seconds": 9722.050307468046, "env_steps_per_second": 4114.358467089373, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7005695298522348, "eta_seconds": 9722.050307468046, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.002776364052704955, "chaser_grad_norm": 0.06105522019788623, "chaser_new_fields": 93.875, "chaser_partner_vision": 0.0, "chaser_return": 8.65999984741211, "chaser_subspace_norm": 53.78749084472656, "collisions_per_episode": 1.0, "entropy": 0.0024248933186754584, "explorer_grad_norm": 0.4683804363012314, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.6875, "explorer_subspace_norm": 97.96976470947266, "final_distance": 2.3905584812164307, "max_param_abs": 2.239759922027588, "policy_loss": -0.0006870093730952931, "value_loss": 0.07492147944867611}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:40:10.225191+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032010.2252047, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.835848784226659, "elapsed_seconds": 10209.055623215972, "env_steps_per_second": 4113.994628894921, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5657679476336437, "eta_seconds": 9236.764611481118, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 5.5746936133201086e-05, "chaser_grad_norm": 0.0356141347438097, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.770000457763672, "chaser_subspace_norm": 55.35414123535156, "collisions_per_episode": 1.024999976158142, "entropy": 0.003112861479166895, "explorer_grad_norm": 0.4594806991517544, "explorer_new_fields": 97.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 100.5293960571289, "final_distance": 2.288546323776245, "max_param_abs": 2.2687058448791504, "policy_loss": -3.996277139961535e-05, "value_loss": 0.09902623109519482}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:48:17.230506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032497.2305174, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.971485036740002, "elapsed_seconds": 10697.346132264007, "env_steps_per_second": 4113.169701716266, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.431215030060002, "eta_seconds": 8752.374108216007, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00013370348195351767, "chaser_grad_norm": 0.020539223216474056, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 54.3487434387207, "collisions_per_episode": 0.8500000238418579, "entropy": 0.003950043930672109, "explorer_grad_norm": 0.31602199375629425, "explorer_new_fields": 97.375, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 107.02349090576172, "final_distance": 2.685612916946411, "max_param_abs": 2.2876455783843994, "policy_loss": -0.00022414303993123497, "value_loss": 0.07805249281227589}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:56:25.521017+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032985.521044, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1068254964366657, "elapsed_seconds": 11184.571787171997, "env_steps_per_second": 4112.808328769378, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.296349279974927, "eta_seconds": 8266.857407909738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": -6.989048116023966e-06, "chaser_grad_norm": 0.03843334084376693, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 54.69615936279297, "collisions_per_episode": 0.949999988079071, "entropy": 0.002638183650560677, "explorer_grad_norm": 0.32337915152311325, "explorer_new_fields": 97.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 104.26318359375, "final_distance": 2.929190158843994, "max_param_abs": 2.3113341331481934, "policy_loss": -0.00012264920040583327, "value_loss": 0.09451063349843025}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:04:32.746670+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033472.746682, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.241158140361949, "elapsed_seconds": 11668.169305303018, "env_steps_per_second": 4113.755872413051, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.160772093574633, "eta_seconds": 7778.779536868678, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0006010586898232972, "chaser_grad_norm": 0.09262969810515642, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.769999504089355, "chaser_subspace_norm": 58.6978759765625, "collisions_per_episode": 0.7750000357627869, "entropy": 0.005693767685443163, "explorer_grad_norm": 0.5575678050518036, "explorer_new_fields": 96.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 94.5999984741211, "explorer_subspace_norm": 107.43771362304688, "final_distance": 2.5675151348114014, "max_param_abs": 2.3327088356018066, "policy_loss": -0.0006506353926347686, "value_loss": 0.14505011588335037}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:12:36.344190+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033956.344204, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.376459176991387, "elapsed_seconds": 12155.253037168994, "env_steps_per_second": 4113.447893442225, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.025875506194832, "eta_seconds": 7293.151822301395, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0007247018451639331, "chaser_grad_norm": 0.03711062157526612, "chaser_new_fields": 93.125, "chaser_partner_vision": 0.0, "chaser_return": 8.550000190734863, "chaser_subspace_norm": 60.891971588134766, "collisions_per_episode": 1.149999976158142, "entropy": 0.003253880888223648, "explorer_grad_norm": 0.3885218985378742, "explorer_new_fields": 97.375, "explorer_partner_vision": 0.0, "explorer_return": 95.38750457763672, "explorer_subspace_norm": 105.75569152832031, "final_distance": 3.1141555309295654, "max_param_abs": 2.343086004257202, "policy_loss": -0.00020295001997094886, "value_loss": 0.10075781308114529}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:20:43.427921+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034443.4279327, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5124197246261044, "elapsed_seconds": 12644.711008653976, "env_steps_per_second": 4112.391336141369, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8913029286448255, "eta_seconds": 6808.690543121372, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.000679097149712593, "chaser_grad_norm": 0.03848004946485162, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 58.41977310180664, "collisions_per_episode": 1.024999976158142, "entropy": 0.003096371015999466, "explorer_grad_norm": 0.3082907684147358, "explorer_new_fields": 97.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 111.32701110839844, "final_distance": 2.35443115234375, "max_param_abs": 2.37701678276062, "policy_loss": -0.0004093637610793621, "value_loss": 0.07662378810346127}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:28:52.885893+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034932.8859081, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6460182049038767, "elapsed_seconds": 13125.665537653957, "env_steps_per_second": 4114.077099183178, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7554902468055704, "eta_seconds": 6319.7648885000535, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -4.818960044783438e-05, "chaser_grad_norm": 0.046644340734928846, "chaser_new_fields": 94.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.780000686645508, "chaser_subspace_norm": 59.496761322021484, "collisions_per_episode": 0.7750000357627869, "entropy": 0.002635331475175917, "explorer_grad_norm": 0.36724893748760223, "explorer_new_fields": 97.30000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 115.20111083984375, "final_distance": 2.6180009841918945, "max_param_abs": 2.387871026992798, "policy_loss": -0.0002843060591501967, "value_loss": 0.06970574520528316}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:36:53.840420+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035413.8404324, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7837413430186118, "elapsed_seconds": 13621.468834867002, "env_steps_per_second": 4111.157223856525, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6216034327222624, "eta_seconds": 5837.772357800144, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -4.5233418649925916e-05, "chaser_grad_norm": 0.031938156113028526, "chaser_new_fields": 94.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.734999656677246, "chaser_subspace_norm": 64.35379791259766, "collisions_per_episode": 0.9000000357627869, "entropy": 0.00153730483725667, "explorer_grad_norm": 0.3889255002140999, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 114.7887191772461, "final_distance": 2.327620267868042, "max_param_abs": 2.4029719829559326, "policy_loss": -2.1636963527527087e-05, "value_loss": 0.07703442126512527}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:45:09.643720+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035909.6437354, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9194211592455392, "elapsed_seconds": 14109.916173283942, "env_steps_per_second": 4110.58430735532, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4866769914379632, "eta_seconds": 5352.037169176668, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0006770159199660508, "chaser_grad_norm": 0.04349134210497141, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 64.44725799560547, "collisions_per_episode": 1.0500000715255737, "entropy": 0.00310589128639549, "explorer_grad_norm": 0.5458345338702202, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 113.97291564941406, "final_distance": 2.190478563308716, "max_param_abs": 2.4299399852752686, "policy_loss": -0.00025005317962478735, "value_loss": 0.11118910647928715}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:53:18.091056+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036398.0910678, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.054119260596118, "elapsed_seconds": 14594.829338146024, "env_steps_per_second": 4111.044987910888, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3513730868653726, "eta_seconds": 4864.943112715341, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.002014808165455406, "chaser_grad_norm": 0.09840812161564827, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.769999504089355, "chaser_subspace_norm": 64.50088500976562, "collisions_per_episode": 1.0, "entropy": 0.003318981733173132, "explorer_grad_norm": 0.2116052582859993, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 119.79007720947266, "final_distance": 2.0650737285614014, "max_param_abs": 2.4542784690856934, "policy_loss": -0.0007690534833804108, "value_loss": 0.1076736319810152}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:01:23.004223+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036883.004239, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.192042195535, "elapsed_seconds": 15091.351903925999, "env_steps_per_second": 4108.313184577637, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2170445083811288, "eta_seconds": 4381.360230172064, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 3.25234731755053e-05, "chaser_grad_norm": 0.0462856306694448, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.805000305175781, "chaser_subspace_norm": 65.58392333984375, "collisions_per_episode": 1.100000023841858, "entropy": 0.0020954415667802095, "explorer_grad_norm": 0.45636197179555893, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 114.12955474853516, "final_distance": 1.9388630390167236, "max_param_abs": 2.4749927520751953, "policy_loss": -0.00025378973339762467, "value_loss": 0.0852540098130703}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:09:39.526788+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037379.526802, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.329286794559448, "elapsed_seconds": 15585.432460414013, "env_steps_per_second": 4106.398726025463, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.082321698639862, "eta_seconds": 3896.3581151035037, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0016604845470581386, "chaser_grad_norm": 0.06046704109758139, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 63.56222915649414, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0028125346871092916, "explorer_grad_norm": 0.4928126856684685, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 120.77127838134766, "final_distance": 2.086970806121826, "max_param_abs": 2.497678518295288, "policy_loss": -0.0012127018691963087, "value_loss": 0.08045056462287903}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:17:53.607346+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037873.6073606, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.462656198671676, "elapsed_seconds": 16065.562315218034, "env_steps_per_second": 4108.166194561505, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9466240421424769, "eta_seconds": 3407.8465517129166, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.017760501632627504, "chaser_grad_norm": 0.05853390693664551, "chaser_new_fields": 95.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.944999694824219, "chaser_subspace_norm": 63.48599624633789, "collisions_per_episode": 0.875, "entropy": 0.002385914354817942, "explorer_grad_norm": 0.3553185537457466, "explorer_new_fields": 97.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.0875015258789, "explorer_subspace_norm": 115.29452514648438, "final_distance": 1.882149577140808, "max_param_abs": 2.5180764198303223, "policy_loss": -9.492970397051437e-05, "value_loss": 0.09502151980996132}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:25:53.737198+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038353.737209, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.599196517355288, "elapsed_seconds": 16557.107462479034, "env_steps_per_second": 4106.997563076674, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8116229148274036, "eta_seconds": 2921.8424933786528, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0009985900814508675, "chaser_grad_norm": 0.08174319146201015, "chaser_new_fields": 95.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.940000534057617, "chaser_subspace_norm": 64.16216278076172, "collisions_per_episode": 0.699999988079071, "entropy": 0.003951519611291587, "explorer_grad_norm": 0.40092281624674797, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 122.30136108398438, "final_distance": 1.778633952140808, "max_param_abs": 2.5919830799102783, "policy_loss": -0.00033798273565466275, "value_loss": 0.053136701695621014}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:05.282348+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038845.2823637, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.73444926007751, "elapsed_seconds": 17044.017336279037, "env_steps_per_second": 4107.012954686541, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6763498942967873, "eta_seconds": 2434.859619468434, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0001943756072434466, "chaser_grad_norm": 0.039517590776085854, "chaser_new_fields": 95.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.9350004196167, "chaser_subspace_norm": 69.97343444824219, "collisions_per_episode": 1.0500000715255737, "entropy": 0.001038317845086567, "explorer_grad_norm": 0.566088929772377, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.6875, "explorer_subspace_norm": 118.8498764038086, "final_distance": 1.9658924341201782, "max_param_abs": 2.7334179878234863, "policy_loss": -0.00018318272277229042, "value_loss": 0.07265921123325825}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:12.192219+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039332.1922312, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.8722572013183445, "elapsed_seconds": 17540.12592474604, "env_steps_per_second": 4104.8736085993905, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5413619112575938, "eta_seconds": 1948.9028805273379, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00040952912746639036, "chaser_grad_norm": 0.06409427057951689, "chaser_new_fields": 94.3499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 68.85417175292969, "collisions_per_episode": 1.0, "entropy": 0.002259656845126301, "explorer_grad_norm": 0.3050736915320158, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 126.66679382324219, "final_distance": 1.6908377408981323, "max_param_abs": 2.8545033931732178, "policy_loss": -0.00034200622106395073, "value_loss": 0.11007564514875412}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:50:28.300808+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039828.3008223, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.005594914789447, "elapsed_seconds": 18020.14169324201, "env_steps_per_second": 4106.515989702333, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40585904714509036, "eta_seconds": 1461.0925697223254, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0029222532060817485, "chaser_grad_norm": 0.10623816214501858, "chaser_new_fields": 94.67500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 70.37963104248047, "collisions_per_episode": 1.0500000715255737, "entropy": 0.002306440786924213, "explorer_grad_norm": 0.5238702520728111, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.20000457763672, "explorer_subspace_norm": 122.44673919677734, "final_distance": 1.840972900390625, "max_param_abs": 2.980407953262329, "policy_loss": -0.0007874917932753078, "value_loss": 0.14899397641420364}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:58:28.316576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040308.3165872, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.1398212427180585, "elapsed_seconds": 18503.35647378501, "env_steps_per_second": 4107.362905085598, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2705169075114767, "eta_seconds": 973.8608670413163, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0019687057694612875, "chaser_grad_norm": 0.11068585515022278, "chaser_new_fields": 94.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.760000228881836, "chaser_subspace_norm": 69.50489807128906, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004220188595354557, "explorer_grad_norm": 0.8308986872434616, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 132.56944274902344, "final_distance": 2.0065629482269287, "max_param_abs": 3.1132256984710693, "policy_loss": -0.0011765843178110913, "value_loss": 0.10864873044192791}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:06:31.531358+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040791.531373, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.273368907361938, "elapsed_seconds": 18984.128066502977, "env_steps_per_second": 4108.695417917511, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13521458736825484, "eta_seconds": 486.7725145257174, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00016270278459473708, "chaser_grad_norm": 0.04523493256419897, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.760001182556152, "chaser_subspace_norm": 71.9196548461914, "collisions_per_episode": 0.800000011920929, "entropy": 0.0027996512362733483, "explorer_grad_norm": 0.46223966032266617, "explorer_new_fields": 97.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 128.7664031982422, "final_distance": 2.235522985458374, "max_param_abs": 3.252021074295044, "policy_loss": -0.00030302145377447687, "value_loss": 0.08791894093155861}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:14:32.302950+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041272.3029628, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.406687021900014, "elapsed_seconds": 19464.07327884005, "env_steps_per_second": 4110.136601621321, "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-transformer/non_social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:22:32.248162+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041752.2481732, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.406713302494997, "elapsed_seconds": 19464.16788898199, "env_steps_per_second": 4110.116623340745, "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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:22:32.342773+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041752.3427885, "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..d2b82690793aba5f1061b2e1912fa4df593edff1 --- /dev/null +++ b/non_social/seed_0006/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.406713302494997, + "elapsed_seconds": 19464.16788898199, + "env_steps_per_second": 4110.116623340745, + "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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 5.701445683692352e-05, + "chaser_grad_norm": 0.033485319931060076, + "chaser_new_fields": 94.57500457763672, + "chaser_partner_vision": 0.0, + "chaser_return": 8.830000877380371, + "chaser_subspace_norm": 70.88212585449219, + "collisions_per_episode": 0.9000000357627869, + "entropy": 0.0015748643199913204, + "explorer_grad_norm": 0.313498567789793, + "explorer_new_fields": 97.55000305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 95.80000305175781, + "explorer_subspace_norm": 132.86404418945312, + "final_distance": 2.430217742919922, + "max_param_abs": 3.379441976547241, + "policy_loss": -4.615307129274271e-05, + "value_loss": 0.07251793146133423 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:22:32.342773+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041752.3427885, + "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..162e5a92ff9bddaac51754344ec20f6a595a9b44 --- /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 sha256:021e19e4992a272e4a09a90a4bb4bbedf2703d0d2a2122af4b17a128246da307 +size 29760368 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..06ed280a54d3026e2df518d1b12d10f725cc0116 --- /dev/null +++ b/non_social/seed_0007/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..0da97cf7aa5962baf75ec6a35f14789b890f7162 --- /dev/null +++ b/non_social/seed_0007/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007321535947266966, "elapsed_seconds": 2.635752941016108, "env_steps_per_second": 1517.5929191823122, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.642339740939205, "eta_seconds": 52712.42306738114, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0056701702655459485, "chaser_grad_norm": 7.788321316242218, "chaser_new_fields": 33.400001525878906, "chaser_partner_vision": 0.0, "chaser_return": -3.3650002479553223, "chaser_subspace_norm": 0.03568364679813385, "collisions_per_episode": 1.6750000715255737, "entropy": 2.754485070705414, "explorer_grad_norm": 2.692690923810005, "explorer_new_fields": 34.22500228881836, "explorer_partner_vision": 0.0, "explorer_return": 0.887499988079071, "explorer_subspace_norm": 0.0367908738553524, "final_distance": 5.377388000488281, "max_param_abs": 1.0012009143829346, "policy_loss": -0.013758847162122834, "value_loss": 8.37043160200119}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:21.024207+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022301.0242202, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13322451510667127, "elapsed_seconds": 479.6082543840166, "env_steps_per_second": 4170.070013846392, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.19575608916018, "eta_seconds": 18704.721920976648, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0010951162351031662, "chaser_grad_norm": 0.20734190195798874, "chaser_new_fields": 88.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 7.635000705718994, "chaser_subspace_norm": 11.38082504272461, "collisions_per_episode": 1.5500000715255737, "entropy": 0.10986384190618992, "explorer_grad_norm": 9.22214961051941, "explorer_new_fields": 90.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 84.70000457763672, "explorer_subspace_norm": 14.508320808410645, "final_distance": 3.848186492919922, "max_param_abs": 1.3093934059143066, "policy_loss": -0.0014662645729117685, "value_loss": 0.5799617916345596}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:06:17.996708+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022777.9967215, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2685581553733209, "elapsed_seconds": 966.8093593439553, "env_steps_per_second": 4137.320311746121, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.102604952093097, "eta_seconds": 18369.37782753515, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004036500377779806, "chaser_grad_norm": 0.3623685669153929, "chaser_new_fields": 89.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 7.950000286102295, "chaser_subspace_norm": 15.89876937866211, "collisions_per_episode": 1.0750000476837158, "entropy": 0.021938718855381012, "explorer_grad_norm": 4.207045912742615, "explorer_new_fields": 93.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 89.80000305175781, "explorer_subspace_norm": 30.057382583618164, "final_distance": 2.738595724105835, "max_param_abs": 1.418306827545166, "policy_loss": -0.0025720210731989823, "value_loss": 0.16101448982954025}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:14:25.197813+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023265.1978273, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.40541097576998353, "elapsed_seconds": 1459.4795127719408, "env_steps_per_second": 4111.054624264235, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.000068701163131, "eta_seconds": 18000.247324187272, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0006706118397294958, "chaser_grad_norm": 0.14210382476449013, "chaser_new_fields": 89.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 7.940001010894775, "chaser_subspace_norm": 20.30768585205078, "collisions_per_episode": 1.350000023841858, "entropy": 0.013704919721931219, "explorer_grad_norm": 3.1649893522262573, "explorer_new_fields": 93.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 89.7249984741211, "explorer_subspace_norm": 38.59990692138672, "final_distance": 2.899866819381714, "max_param_abs": 1.4806946516036987, "policy_loss": -0.0010154848629611024, "value_loss": 0.22510406747460365}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:37.867964+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023757.8679757, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5388360899216624, "elapsed_seconds": 1939.809923717985, "env_steps_per_second": 4124.115410579301, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.849524809294962, "eta_seconds": 17458.289313461864, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0004225196101306228, "chaser_grad_norm": 0.12929856218397617, "chaser_new_fields": 90.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.105000495910645, "chaser_subspace_norm": 24.430145263671875, "collisions_per_episode": 0.42500001192092896, "entropy": 0.018239749129861593, "explorer_grad_norm": 2.4173524081707, "explorer_new_fields": 94.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.375, "explorer_subspace_norm": 48.050537109375, "final_distance": 3.496685743331909, "max_param_abs": 1.5447101593017578, "policy_loss": -0.0017914791642823502, "value_loss": 0.08978062123060226}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:38.198377+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024238.198392, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6757249187127713, "elapsed_seconds": 2432.6097073659766, "env_steps_per_second": 4110.811516421997, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.730074430989398, "eta_seconds": 17028.267951561833, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0002855049126340506, "chaser_grad_norm": 0.04755624569952488, "chaser_new_fields": 92.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 26.357845306396484, "collisions_per_episode": 0.4749999940395355, "entropy": 0.00948554091155529, "explorer_grad_norm": 1.8195624947547913, "explorer_new_fields": 94.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 56.789939880371094, "final_distance": 3.4260756969451904, "max_param_abs": 1.5881214141845703, "policy_loss": -0.0005307350391525745, "value_loss": 0.07191452570259571}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:50.998161+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024730.9981747, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.812102520614151, "elapsed_seconds": 2923.5690742109437, "env_steps_per_second": 4104.572081382664, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.601914283480189, "eta_seconds": 16566.891420528682, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0006739896190994799, "chaser_grad_norm": 0.10493527818471193, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.385001182556152, "chaser_subspace_norm": 29.47441291809082, "collisions_per_episode": 0.6000000238418579, "entropy": 0.006973615032620728, "explorer_grad_norm": 0.6466992497444153, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.75, "explorer_subspace_norm": 63.80134963989258, "final_distance": 3.5140304565429688, "max_param_abs": 1.62350332736969, "policy_loss": -0.0006643085562352269, "value_loss": 0.07135201618075371}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:47:01.957527+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025221.9575417, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9456151059741711, "elapsed_seconds": 3404.214381507016, "env_steps_per_second": 4112.549455185112, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.4578997853068065, "eta_seconds": 16048.439227104504, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0005603134956231881, "chaser_grad_norm": 0.14965984225273132, "chaser_new_fields": 91.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.290000915527344, "chaser_subspace_norm": 29.964725494384766, "collisions_per_episode": 0.675000011920929, "entropy": 0.006169264786876738, "explorer_grad_norm": 1.037157267332077, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 70.84859466552734, "final_distance": 2.876924514770508, "max_param_abs": 1.64976966381073, "policy_loss": -0.0005626203031274812, "value_loss": 0.08296688832342625}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:55:02.602833+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025702.6028457, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0822247666455547, "elapsed_seconds": 3896.009159923997, "env_steps_per_second": 4106.766525238901, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.328899066582219, "eta_seconds": 15584.036639695987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0003650248149850155, "chaser_grad_norm": 0.11409659311175346, "chaser_new_fields": 91.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.190001487731934, "chaser_subspace_norm": 30.466596603393555, "collisions_per_episode": 0.6000000238418579, "entropy": 0.005053916946053505, "explorer_grad_norm": 1.1245538145303726, "explorer_new_fields": 94.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 73.21965026855469, "final_distance": 3.1280291080474854, "max_param_abs": 1.6762510538101196, "policy_loss": -0.0008933144271168203, "value_loss": 0.08213349059224129}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:03:14.397612+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026194.397625, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2177029859505517, "elapsed_seconds": 4383.730749421986, "env_steps_per_second": 4106.091598434366, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.194310284940789, "eta_seconds": 15099.51702578684, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 4.095516214785988e-05, "chaser_grad_norm": 0.025947158224880695, "chaser_new_fields": 91.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.270000457763672, "chaser_subspace_norm": 32.30112838745117, "collisions_per_episode": 0.45000001788139343, "entropy": 0.0026165806339122355, "explorer_grad_norm": 0.36463233456015587, "explorer_new_fields": 94.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 79.55625915527344, "final_distance": 3.1705706119537354, "max_param_abs": 1.703834056854248, "policy_loss": -3.1307698380000915e-05, "value_loss": 0.030478816013783216}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:11:22.119202+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026682.119213, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3514639576638647, "elapsed_seconds": 4865.270247589913, "env_steps_per_second": 4110.76856622863, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.054391872991594, "eta_seconds": 14595.81074276974, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0001641448992101835, "chaser_grad_norm": 0.21693343855440617, "chaser_new_fields": 91.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.26500129699707, "chaser_subspace_norm": 36.309104919433594, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0043490296229720116, "explorer_grad_norm": 0.9580365717411041, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 85.16812133789062, "final_distance": 2.5449771881103516, "max_param_abs": 1.7264920473098755, "policy_loss": -0.00027526332845528856, "value_loss": 0.06534147821366787}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:23.658701+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027163.658716, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4883823616383274, "elapsed_seconds": 5358.176501897979, "env_steps_per_second": 4105.874450423038, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9239171352283173, "eta_seconds": 14126.101686821943, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00028120612496956854, "chaser_grad_norm": 0.10291431844234467, "chaser_new_fields": 91.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.290000915527344, "chaser_subspace_norm": 36.331050872802734, "collisions_per_episode": 0.4000000059604645, "entropy": 0.003830882895272225, "explorer_grad_norm": 0.43580418080091476, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.2750015258789, "explorer_subspace_norm": 88.15481567382812, "final_distance": 3.4332566261291504, "max_param_abs": 1.7417998313903809, "policy_loss": -0.0003201260770566705, "value_loss": 0.03968103416264057}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:36.564954+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027656.564966, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6262635646850039, "elapsed_seconds": 5854.548832866014, "env_steps_per_second": 4099.376516473794, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.794614984265009, "eta_seconds": 13660.613943354034, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.001304308083141692, "chaser_grad_norm": 0.12768317200243473, "chaser_new_fields": 91.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.200000762939453, "chaser_subspace_norm": 38.97804260253906, "collisions_per_episode": 0.6000000238418579, "entropy": 0.0038849215488880873, "explorer_grad_norm": 0.6264419853687286, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.05000305175781, "explorer_subspace_norm": 92.731689453125, "final_distance": 3.383495330810547, "max_param_abs": 1.7643274068832397, "policy_loss": -0.0009467237654607885, "value_loss": 0.1024355310946703}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:52.937286+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028152.9373002, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.761267324278047, "elapsed_seconds": 6340.562367400969, "env_steps_per_second": 4100.582644478827, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.658016750423636, "eta_seconds": 13168.86030152509, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.02061704900578669, "chaser_grad_norm": 0.06211651675403118, "chaser_new_fields": 91.5, "chaser_partner_vision": 0.0, "chaser_return": 8.260001182556152, "chaser_subspace_norm": 39.97160720825195, "collisions_per_episode": 1.024999976158142, "entropy": 0.005657536443322897, "explorer_grad_norm": 0.5513572990894318, "explorer_new_fields": 93.20000457763672, "explorer_partner_vision": 0.0, "explorer_return": 89.2750015258789, "explorer_subspace_norm": 97.76179504394531, "final_distance": 3.0771918296813965, "max_param_abs": 1.7837369441986084, "policy_loss": -0.007811741366051006, "value_loss": 0.2216559238731861}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:58.950819+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028638.9508302, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8960263306055438, "elapsed_seconds": 6825.694790179958, "env_steps_per_second": 4102.1465009369085, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.521191756838867, "eta_seconds": 12676.290324619922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00010724438518250334, "chaser_grad_norm": 0.05691736377775669, "chaser_new_fields": 91.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.345000267028809, "chaser_subspace_norm": 42.32274627685547, "collisions_per_episode": 0.5, "entropy": 0.0019734848174266517, "explorer_grad_norm": 1.031820498406887, "explorer_new_fields": 95.2249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 97.07659912109375, "final_distance": 3.2742631435394287, "max_param_abs": 1.7945631742477417, "policy_loss": -9.86337696086137e-05, "value_loss": 0.03548165410757065}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:52:04.083243+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029124.0832567, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.029998636115536, "elapsed_seconds": 7307.995090015931, "env_steps_per_second": 4105.093070052213, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3833310601925612, "eta_seconds": 12179.99181669322, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0005490491454267499, "chaser_grad_norm": 0.08179247425869107, "chaser_new_fields": 91.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.335000991821289, "chaser_subspace_norm": 44.21648406982422, "collisions_per_episode": 0.550000011920929, "entropy": 0.0029834862798452377, "explorer_grad_norm": 0.26826261729002, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 104.014892578125, "final_distance": 2.916912078857422, "max_param_abs": 1.8152722120285034, "policy_loss": -0.0008010073231989878, "value_loss": 0.03606746532022953}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:00:06.383542+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029606.3835535, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1659557833094407, "elapsed_seconds": 7797.440819913987, "env_steps_per_second": 4103.9105956757985, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.248933674964162, "eta_seconds": 11696.161229870982, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": -0.00014074789685480926, "chaser_grad_norm": 0.027934850193560123, "chaser_new_fields": 92.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.36500072479248, "chaser_subspace_norm": 44.905941009521484, "collisions_per_episode": 0.6500000357627869, "entropy": 0.003372886450961232, "explorer_grad_norm": 0.555730439722538, "explorer_new_fields": 95.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 102.32695770263672, "final_distance": 3.1198642253875732, "max_param_abs": 1.8329010009765625, "policy_loss": -0.000740330724617877, "value_loss": 0.08236990869045258}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:08:15.829273+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030095.829288, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.30484279540998, "elapsed_seconds": 8297.434063475928, "env_steps_per_second": 4097.652327201122, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.118316723201738, "eta_seconds": 11225.940203526257, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.009740977395011674, "chaser_grad_norm": 0.6714459452778101, "chaser_new_fields": 91.5, "chaser_partner_vision": 0.0, "chaser_return": 8.26500129699707, "chaser_subspace_norm": 44.96942138671875, "collisions_per_episode": 0.4749999940395355, "entropy": 0.004150857916101813, "explorer_grad_norm": 0.3438621796667576, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.5374984741211, "explorer_subspace_norm": 104.07538604736328, "final_distance": 3.0234169960021973, "max_param_abs": 1.843275547027588, "policy_loss": -0.0036788812899146706, "value_loss": 0.04113639518618584}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:16:35.822515+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030595.8225272, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4409293631374993, "elapsed_seconds": 8787.345707294997, "env_steps_per_second": 4096.800239703083, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9833581105013876, "eta_seconds": 10740.089197804995, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00032544704121972856, "chaser_grad_norm": 0.059189712163060904, "chaser_new_fields": 91.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.315000534057617, "chaser_subspace_norm": 47.63878631591797, "collisions_per_episode": 0.625, "entropy": 0.0025119376950897276, "explorer_grad_norm": 0.1984409336000681, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 107.43093872070312, "final_distance": 3.101303815841675, "max_param_abs": 1.8630828857421875, "policy_loss": -0.0003497038057531654, "value_loss": 0.05271161999553442}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:45.734160+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031085.7341752, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.578201281351115, "elapsed_seconds": 9281.524612864014, "env_steps_per_second": 4094.154956754921, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8495908899143902, "eta_seconds": 10258.527203691805, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.02480472929728439, "chaser_grad_norm": 0.244842697866261, "chaser_new_fields": 91.625, "chaser_partner_vision": 0.0, "chaser_return": 8.315001487731934, "chaser_subspace_norm": 51.06640625, "collisions_per_episode": 0.550000011920929, "entropy": 0.003615096095018089, "explorer_grad_norm": 0.7532136589288712, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.0875015258789, "explorer_subspace_norm": 110.84048461914062, "final_distance": 3.354170322418213, "max_param_abs": 1.8806450366973877, "policy_loss": 0.0008579917270594706, "value_loss": 0.05694609973579645}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:59.913065+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031579.9130762, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.714499016109427, "elapsed_seconds": 9772.196457993938, "env_steps_per_second": 4093.2455842390327, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.714499016109427, "eta_seconds": 9772.196457993938, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00027975325311802024, "chaser_grad_norm": 0.0888599343597889, "chaser_new_fields": 91.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.335000038146973, "chaser_subspace_norm": 51.72610092163086, "collisions_per_episode": 0.574999988079071, "entropy": 0.0048523551085963845, "explorer_grad_norm": 0.48606380820274353, "explorer_new_fields": 95.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 113.36235809326172, "final_distance": 2.3708882331848145, "max_param_abs": 1.8992223739624023, "policy_loss": -0.0006996984827565278, "value_loss": 0.07033180631697178}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:41:10.584911+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032070.584927, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.849816345538032, "elapsed_seconds": 10259.338843936916, "env_steps_per_second": 4093.83105859899, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5784052650106, "eta_seconds": 9282.258954038161, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0010424142872582642, "chaser_grad_norm": 0.046187774278223515, "chaser_new_fields": 91.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.350000381469727, "chaser_subspace_norm": 53.15707778930664, "collisions_per_episode": 0.7750000357627869, "entropy": 0.004856990068219602, "explorer_grad_norm": 0.49370693415403366, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.1624984741211, "explorer_subspace_norm": 112.80384826660156, "final_distance": 2.9221484661102295, "max_param_abs": 1.9079666137695312, "policy_loss": -0.0013670023337475357, "value_loss": 0.1550607569515705}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:49:17.727297+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032557.7273114, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.985300520084145, "elapsed_seconds": 10747.081872302922, "env_steps_per_second": 4094.1346239666755, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4425186073415732, "eta_seconds": 8793.066986429663, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0006623503939650055, "chaser_grad_norm": 0.02257010107859969, "chaser_new_fields": 92.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.430000305175781, "chaser_subspace_norm": 55.4481201171875, "collisions_per_episode": 0.4749999940395355, "entropy": 0.0018681079091038555, "explorer_grad_norm": 0.37577641010284424, "explorer_new_fields": 95.625, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 114.24705505371094, "final_distance": 2.8613860607147217, "max_param_abs": 1.9207593202590942, "policy_loss": -0.0005284996259726782, "value_loss": 0.029496636241674423}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:25.470326+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033045.4703398, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1197505763033404, "elapsed_seconds": 11231.102074692026, "env_steps_per_second": 4095.7690255220473, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.305902599876382, "eta_seconds": 8301.249359554975, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0022927334245759212, "chaser_grad_norm": 0.08188355341553688, "chaser_new_fields": 91.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.255000114440918, "chaser_subspace_norm": 54.03846740722656, "collisions_per_episode": 0.9000000357627869, "entropy": 0.004221938259433955, "explorer_grad_norm": 0.6325102597475052, "explorer_new_fields": 94.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.1624984741211, "explorer_subspace_norm": 122.81529235839844, "final_distance": 3.04292893409729, "max_param_abs": 1.9428632259368896, "policy_loss": -0.0018103447958992547, "value_loss": 0.15741181001067162}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:29.490526+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033529.490538, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2574594580505316, "elapsed_seconds": 11726.854048981913, "env_steps_per_second": 4093.169387075914, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1716396387003543, "eta_seconds": 7817.902699321276, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0005912225335436005, "chaser_grad_norm": 0.11593510583043098, "chaser_new_fields": 92.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.410000801086426, "chaser_subspace_norm": 55.07162094116211, "collisions_per_episode": 0.42500001192092896, "entropy": 0.002368191140703857, "explorer_grad_norm": 0.35706258565187454, "explorer_new_fields": 95.9000015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.55000305175781, "explorer_subspace_norm": 116.91510772705078, "final_distance": 2.7921955585479736, "max_param_abs": 1.9471858739852905, "policy_loss": -0.0002902300445128958, "value_loss": 0.05704915523529053}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:13:45.242502+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034025.2425165, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3954701729469363, "elapsed_seconds": 12223.69262260897, "env_steps_per_second": 4090.4169913042383, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.037282103768162, "eta_seconds": 7334.215573565382, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0006532528753778605, "chaser_grad_norm": 0.018125110305845737, "chaser_new_fields": 92.375, "chaser_partner_vision": 0.0, "chaser_return": 8.460000991821289, "chaser_subspace_norm": 57.62477493286133, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0014427294954657555, "explorer_grad_norm": 0.22150558233261108, "explorer_new_fields": 95.875, "explorer_partner_vision": 0.0, "explorer_return": 93.5875015258789, "explorer_subspace_norm": 119.5532455444336, "final_distance": 3.2720417976379395, "max_param_abs": 1.965879201889038, "policy_loss": -0.00031460524752002783, "value_loss": 0.02565121604129672}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:22:02.081076+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034522.0810897, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5311831382352588, "elapsed_seconds": 12712.259297646931, "env_steps_per_second": 4090.53959508404, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.901406305203601, "eta_seconds": 6845.062698732963, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 2.916654861073731e-05, "chaser_grad_norm": 0.022575194481760263, "chaser_new_fields": 92.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.370001792907715, "chaser_subspace_norm": 58.1812858581543, "collisions_per_episode": 0.42500001192092896, "entropy": 0.001182295469334349, "explorer_grad_norm": 0.17028779909014702, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 125.23107147216797, "final_distance": 2.833322286605835, "max_param_abs": 1.9831078052520752, "policy_loss": 2.6350021962606718e-05, "value_loss": 0.03482671733945608}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:30:10.647752+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035010.6477683, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.669250627363055, "elapsed_seconds": 13209.302258507, "env_steps_per_second": 4088.028189772336, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7666762279896193, "eta_seconds": 6360.034420762629, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -0.00010172467695129805, "chaser_grad_norm": 0.029083694331347942, "chaser_new_fields": 92.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.39000129699707, "chaser_subspace_norm": 61.873416900634766, "collisions_per_episode": 0.574999988079071, "entropy": 0.0019262868736404926, "explorer_grad_norm": 0.16040929034352303, "explorer_new_fields": 95.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.9124984741211, "explorer_subspace_norm": 127.52452850341797, "final_distance": 3.283766508102417, "max_param_abs": 1.994425892829895, "policy_loss": -0.00013919115602334386, "value_loss": 0.05453700013458729}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:38:27.690710+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035507.690722, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.804730262123889, "elapsed_seconds": 13697.028943646, "env_steps_per_second": 4088.477890380613, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.630598683767381, "eta_seconds": 5870.1552615625715, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00045638229906344585, "chaser_grad_norm": 0.04505558591336012, "chaser_new_fields": 92.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.405000686645508, "chaser_subspace_norm": 60.650604248046875, "collisions_per_episode": 0.42500001192092896, "entropy": 0.0021596268052235246, "explorer_grad_norm": 0.18482574075460434, "explorer_new_fields": 95.5, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 130.99581909179688, "final_distance": 3.2085633277893066, "max_param_abs": 2.0095036029815674, "policy_loss": -0.0005444307624102862, "value_loss": 0.0479189557954669}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:46:35.417397+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035995.4174125, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.940365896745546, "elapsed_seconds": 14185.317228283966, "env_steps_per_second": 4088.7347858780604, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.494621547041414, "eta_seconds": 5380.63756934909, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 7.796954645455934e-05, "chaser_grad_norm": 0.029224731028079987, "chaser_new_fields": 92.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.410000801086426, "chaser_subspace_norm": 61.1788444519043, "collisions_per_episode": 0.5, "entropy": 0.0019386539643164724, "explorer_grad_norm": 0.16375256329774857, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.32500457763672, "explorer_subspace_norm": 131.7733154296875, "final_distance": 3.331510305404663, "max_param_abs": 2.0207414627075195, "policy_loss": -0.0003814013245291292, "value_loss": 0.03302660305052996}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:54:43.705683+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036483.7056966, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.0740254152475, "elapsed_seconds": 14666.491494891001, "env_steps_per_second": 4090.9579513887625, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3580084717491667, "eta_seconds": 4888.830498297, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 2.8710134759246664e-05, "chaser_grad_norm": 0.06163962371647358, "chaser_new_fields": 92.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.370000839233398, "chaser_subspace_norm": 60.25822067260742, "collisions_per_episode": 0.75, "entropy": 0.0015792029444128275, "explorer_grad_norm": 0.5091016665101051, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 137.88807678222656, "final_distance": 3.1891462802886963, "max_param_abs": 2.0335261821746826, "policy_loss": -5.503750017643583e-05, "value_loss": 0.09783779457211494}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:02:44.879948+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036964.8799622, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.211766529317505, "elapsed_seconds": 15162.35950554302, "env_steps_per_second": 4089.0733383108472, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2227709278663723, "eta_seconds": 4401.975340318941, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0009778332349990038, "chaser_grad_norm": 0.03449817607179284, "chaser_new_fields": 91.75, "chaser_partner_vision": 0.0, "chaser_return": 8.305000305175781, "chaser_subspace_norm": 61.31635665893555, "collisions_per_episode": 0.5, "entropy": 0.002756824775133282, "explorer_grad_norm": 0.15681877918541431, "explorer_new_fields": 95.625, "explorer_partner_vision": 0.0, "explorer_return": 93.13750457763672, "explorer_subspace_norm": 134.85183715820312, "final_distance": 3.1077980995178223, "max_param_abs": 2.056541919708252, "policy_loss": -0.00045683529019413527, "value_loss": 0.033283534459769726}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:11:00.747957+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037460.747969, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.352765393988326, "elapsed_seconds": 15669.955418357975, "env_steps_per_second": 4084.249016115353, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0881913484970815, "eta_seconds": 3917.4888545894937, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00022374950116521575, "chaser_grad_norm": 0.016273183282464743, "chaser_new_fields": 91.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.369999885559082, "chaser_subspace_norm": 62.595252990722656, "collisions_per_episode": 0.30000001192092896, "entropy": 0.0018401675042696297, "explorer_grad_norm": 0.45136695355176926, "explorer_new_fields": 95.92500305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 137.06690979003906, "final_distance": 3.356236219406128, "max_param_abs": 2.076568841934204, "policy_loss": -0.00011893511821114977, "value_loss": 0.033938806504011154}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:19:28.343874+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037968.3438883, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.491672043390831, "elapsed_seconds": 16170.01935620699, "env_steps_per_second": 4081.6277671718044, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9527789182950247, "eta_seconds": 3430.004105862089, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 2.072192314472865e-05, "chaser_grad_norm": 0.027606753166764975, "chaser_new_fields": 92.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.40000057220459, "chaser_subspace_norm": 64.06857299804688, "collisions_per_episode": 0.25, "entropy": 0.0010840338945854455, "explorer_grad_norm": 0.25276608020067215, "explorer_new_fields": 95.8499984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.625, "explorer_subspace_norm": 140.53814697265625, "final_distance": 3.007100820541382, "max_param_abs": 2.088221311569214, "policy_loss": -8.838173859437859e-07, "value_loss": 0.029558385722339153}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:27:48.407810+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038468.4078233, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.627012663554698, "elapsed_seconds": 16657.24558879691, "env_steps_per_second": 4082.3075842583753, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8165316465096526, "eta_seconds": 2939.5139274347493, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00021099224737164413, "chaser_grad_norm": 0.022988375276327133, "chaser_new_fields": 92.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.395000457763672, "chaser_subspace_norm": 65.70854949951172, "collisions_per_episode": 0.3499999940395355, "entropy": 0.001629533973755315, "explorer_grad_norm": 0.15016156248748302, "explorer_new_fields": 95.80000305175781, "explorer_partner_vision": 0.0, "explorer_return": 93.4749984741211, "explorer_subspace_norm": 140.18968200683594, "final_distance": 3.5926601886749268, "max_param_abs": 2.098592758178711, "policy_loss": -0.00015527392574576915, "value_loss": 0.030302355997264385}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:35:55.634043+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038955.634058, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.760307500451939, "elapsed_seconds": 17137.10700162698, "env_steps_per_second": 4084.70344459857, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6800439286359913, "eta_seconds": 2448.1581430895685, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 3.3315275531742294e-05, "chaser_grad_norm": 0.01907614665105939, "chaser_new_fields": 92.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.385001182556152, "chaser_subspace_norm": 67.7673110961914, "collisions_per_episode": 0.32500001788139343, "entropy": 0.0011569378548301756, "explorer_grad_norm": 0.30077899992465973, "explorer_new_fields": 95.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.7750015258789, "explorer_subspace_norm": 137.2908172607422, "final_distance": 2.960758924484253, "max_param_abs": 2.111963987350464, "policy_loss": -0.0001201670193040627, "value_loss": 0.06301439460366964}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:43:55.495453+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039435.495464, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.8964423464341476, "elapsed_seconds": 17627.19244716293, "env_steps_per_second": 4084.598282784862, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5440491496037942, "eta_seconds": 1958.5769385736592, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": -4.707308586138215e-05, "chaser_grad_norm": 0.026942245662212372, "chaser_new_fields": 92.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.40000057220459, "chaser_subspace_norm": 69.9012451171875, "collisions_per_episode": 0.6000000238418579, "entropy": 0.0015254993049893528, "explorer_grad_norm": 0.14233427867293358, "explorer_new_fields": 95.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 143.6483612060547, "final_distance": 3.1813814640045166, "max_param_abs": 2.1261916160583496, "policy_loss": -0.0001265478190859426, "value_loss": 0.037048451602458954}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:52:05.580899+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039925.5809147, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.036134482898601, "elapsed_seconds": 18130.084138434962, "env_steps_per_second": 4081.6137109437527, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4083352283431298, "eta_seconds": 1470.0068220352673, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0009871368235412348, "chaser_grad_norm": 0.19112883135676384, "chaser_new_fields": 90.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 7.960000514984131, "chaser_subspace_norm": 72.46993255615234, "collisions_per_episode": 1.274999976158142, "entropy": 0.006187069229781628, "explorer_grad_norm": 0.29792730137705803, "explorer_new_fields": 94.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 91.67500305175781, "explorer_subspace_norm": 144.86412048339844, "final_distance": 2.9011800289154053, "max_param_abs": 2.1391761302948, "policy_loss": -0.0014070749646037939, "value_loss": 0.22160150110721588}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:00:28.472591+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040428.4726017, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.172838899154158, "elapsed_seconds": 18622.22003695497, "env_steps_per_second": 4081.146063636955, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2722546789028504, "eta_seconds": 980.1168440502614, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00012489964710223067, "chaser_grad_norm": 0.02417089673690498, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.345001220703125, "chaser_subspace_norm": 73.57506561279297, "collisions_per_episode": 0.6000000238418579, "entropy": 0.001758875121595338, "explorer_grad_norm": 0.4114908054471016, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.5374984741211, "explorer_subspace_norm": 150.04708862304688, "final_distance": 2.867208242416382, "max_param_abs": 2.1569149494171143, "policy_loss": 2.0559430755628938e-05, "value_loss": 0.06918449513614178}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:08:40.608490+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040920.6085043, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.307529441521943, "elapsed_seconds": 19107.105989478994, "env_steps_per_second": 4082.2508674494916, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13609049850056265, "eta_seconds": 489.9257946020255, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -1.0290373633425887e-05, "chaser_grad_norm": 0.028074567671865225, "chaser_new_fields": 91.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.36500072479248, "chaser_subspace_norm": 71.74186706542969, "collisions_per_episode": 0.5, "entropy": 0.0018037223198916763, "explorer_grad_norm": 0.18616386130452156, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.13750457763672, "explorer_subspace_norm": 150.32296752929688, "final_distance": 2.9552972316741943, "max_param_abs": 2.166377305984497, "policy_loss": -0.00015115833899415065, "value_loss": 0.054916064254939556}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:16:45.494442+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041405.4944534, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.424764062891094, "elapsed_seconds": 19529.150626407936, "env_steps_per_second": 4096.440317881591, "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-transformer/non_social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:47.539078+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041827.5390878, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.424772208754698, "elapsed_seconds": 19529.17995151691, "env_steps_per_second": 4096.4341666474365, "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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:47.568405+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041827.5684185, "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..321ddf65b5322df71e0537e2656076b9f704e7e1 --- /dev/null +++ b/non_social/seed_0007/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.424772208754698, + "elapsed_seconds": 19529.17995151691, + "env_steps_per_second": 4096.4341666474365, + "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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 5.98452016989659e-05, + "chaser_grad_norm": 0.023093817057088017, + "chaser_new_fields": 91.82500457763672, + "chaser_partner_vision": 0.0, + "chaser_return": 8.360000610351562, + "chaser_subspace_norm": 73.04069519042969, + "collisions_per_episode": 0.3499999940395355, + "entropy": 0.0009539475431665778, + "explorer_grad_norm": 0.2525465562939644, + "explorer_new_fields": 95.95000457763672, + "explorer_partner_vision": 0.0, + "explorer_return": 93.6624984741211, + "explorer_subspace_norm": 149.63381958007812, + "final_distance": 2.9168975353240967, + "max_param_abs": 2.185072422027588, + "policy_loss": -7.590103632537648e-05, + "value_loss": 0.040869749151170254 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:47.568405+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041827.5684185, + "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..061781ad2d7bff9377d6ffd279bb215ade3b8dc0 --- /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 sha256:8bd159d34635da80bf0bdbc948924eb30df6571f77a8dc53b6d138adad560333 +size 29760336 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..3bb22d6def868543e97e049254871de8bec1338c --- /dev/null +++ b/non_social/seed_0008/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..2fbd90a4ca305ba14ce9607228abbbcaeb2588ee --- /dev/null +++ b/non_social/seed_0008/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007065588472242882, "elapsed_seconds": 2.543611850007437, "env_steps_per_second": 1572.5669779327002, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.130470385638537, "eta_seconds": 50869.693388298736, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004931495480879455, "chaser_grad_norm": 5.797280520200729, "chaser_new_fields": 34.400001525878906, "chaser_partner_vision": 0.0, "chaser_return": -3.140000343322754, "chaser_subspace_norm": 0.04043695330619812, "collisions_per_episode": 0.875, "entropy": 2.7503649592399597, "explorer_grad_norm": 4.899797976016998, "explorer_new_fields": 32.67499923706055, "explorer_partner_vision": 0.0, "explorer_return": -1.1750000715255737, "explorer_subspace_norm": 0.03649904206395149, "final_distance": 6.141808986663818, "max_param_abs": 1.001198410987854, "policy_loss": -0.014888535752888288, "value_loss": 8.289218425750732}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:50.773592+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022330.7736073, "update": 1} +{"device": "cuda", "elapsed_hours": 0.1347328106375062, "elapsed_seconds": 485.0381182950223, "env_steps_per_second": 4123.387264964418, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.254579614862742, "eta_seconds": 18916.48661350587, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.00038829119459593553, "chaser_grad_norm": 0.3615344353020191, "chaser_new_fields": 91.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.295001029968262, "chaser_subspace_norm": 10.3538179397583, "collisions_per_episode": 0.550000011920929, "entropy": 0.07660105265676975, "explorer_grad_norm": 11.175984263420105, "explorer_new_fields": 92.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 88.2249984741211, "explorer_subspace_norm": 15.46786117553711, "final_distance": 2.3563103675842285, "max_param_abs": 1.3020983934402466, "policy_loss": -0.001913280138751361, "value_loss": 0.5714928358793259}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:06:53.268103+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022813.268116, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27140319160418586, "elapsed_seconds": 977.0514897750691, "env_steps_per_second": 4093.9500546904196, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.156660640479531, "eta_seconds": 18563.978305726312, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.002967574265182016, "chaser_grad_norm": 0.3910607472062111, "chaser_new_fields": 92.07500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 13.558795928955078, "collisions_per_episode": 0.6000000238418579, "entropy": 0.018904320430010557, "explorer_grad_norm": 4.862184643745422, "explorer_new_fields": 96.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.88750457763672, "explorer_subspace_norm": 29.38630485534668, "final_distance": 2.6767866611480713, "max_param_abs": 1.4646317958831787, "policy_loss": -0.0019092789411061517, "value_loss": 0.20366156846284866}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:15:05.281472+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023305.281485, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4058616884000043, "elapsed_seconds": 1461.1020782400155, "env_steps_per_second": 4106.489265436784, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.005627490266719, "eta_seconds": 18020.25896496019, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00046268375189684186, "chaser_grad_norm": 0.12948780693113804, "chaser_new_fields": 92.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.415000915527344, "chaser_subspace_norm": 16.028596878051758, "collisions_per_episode": 1.1750000715255737, "entropy": 0.00979757378809154, "explorer_grad_norm": 3.904834270477295, "explorer_new_fields": 96.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.2249984741211, "explorer_subspace_norm": 39.12043762207031, "final_distance": 2.462959051132202, "max_param_abs": 1.5578569173812866, "policy_loss": -0.0006532135430017405, "value_loss": 0.10980244167149067}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:23:09.332059+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023789.3320696, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5432946273525078, "elapsed_seconds": 1955.8606584690278, "env_steps_per_second": 4090.2709328291776, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.889651646172569, "eta_seconds": 17602.74592622125, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0017532067771574011, "chaser_grad_norm": 0.17474452033638954, "chaser_new_fields": 92.7750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.505000114440918, "chaser_subspace_norm": 18.587614059448242, "collisions_per_episode": 0.6500000357627869, "entropy": 0.010672278702259064, "explorer_grad_norm": 3.272294759750366, "explorer_new_fields": 97.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.2750015258789, "explorer_subspace_norm": 48.02800369262695, "final_distance": 2.54636549949646, "max_param_abs": 1.6495580673217773, "policy_loss": -0.001415628528683044, "value_loss": 0.12378275021910667}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:31:24.090640+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024284.0906553, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6767533174266767, "elapsed_seconds": 2436.311942736036, "env_steps_per_second": 4104.564700680226, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.737273221986737, "eta_seconds": 17054.183599152253, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00018668962401341416, "chaser_grad_norm": 0.03761191060766578, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.699999809265137, "chaser_subspace_norm": 21.434608459472656, "collisions_per_episode": 0.574999988079071, "entropy": 0.007814222713932395, "explorer_grad_norm": 3.504342168569565, "explorer_new_fields": 97.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.3125, "explorer_subspace_norm": 52.27217483520508, "final_distance": 2.718687057495117, "max_param_abs": 1.7167993783950806, "policy_loss": -0.0003310547100277006, "value_loss": 0.1276181936264038}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:39:24.541922+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024764.5419335, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8104041823064068, "elapsed_seconds": 2917.4550563030643, "env_steps_per_second": 4113.173902739101, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.592290366402971, "eta_seconds": 16532.245319050697, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.003955254434943267, "chaser_grad_norm": 0.12333505228161812, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.61500072479248, "chaser_subspace_norm": 22.41366195678711, "collisions_per_episode": 1.25, "entropy": 0.01306794909760356, "explorer_grad_norm": 1.2877201437950134, "explorer_new_fields": 95.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.7249984741211, "explorer_subspace_norm": 55.14224624633789, "final_distance": 2.611316442489624, "max_param_abs": 1.777976155281067, "policy_loss": -0.003405990671367931, "value_loss": 0.2729642912745476}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:47:25.685038+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025245.6850526, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9461799943163836, "elapsed_seconds": 3406.247979538981, "env_steps_per_second": 4110.094181074518, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.460562830348666, "eta_seconds": 16058.026189255197, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0010479668061823821, "chaser_grad_norm": 0.047439150512218475, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 24.042903900146484, "collisions_per_episode": 0.6000000238418579, "entropy": 0.006697462871670723, "explorer_grad_norm": 1.8050414025783539, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 61.79494094848633, "final_distance": 2.2576606273651123, "max_param_abs": 1.8457244634628296, "policy_loss": -0.00038730431447220326, "value_loss": 0.06200123019516468}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:55:34.477960+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025734.477975, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.081819242059169, "elapsed_seconds": 3894.5492714130087, "env_steps_per_second": 4108.305964298387, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.327276968236676, "eta_seconds": 15578.197085652033, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0036461019300480096, "chaser_grad_norm": 0.1597575843334198, "chaser_new_fields": 93.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.570000648498535, "chaser_subspace_norm": 25.02796745300293, "collisions_per_episode": 1.1750000715255737, "entropy": 0.008321429369971156, "explorer_grad_norm": 1.6470242142677307, "explorer_new_fields": 96.875, "explorer_partner_vision": 0.0, "explorer_return": 94.5250015258789, "explorer_subspace_norm": 67.9875259399414, "final_distance": 2.4514002799987793, "max_param_abs": 1.901628851890564, "policy_loss": -0.002474798800424338, "value_loss": 0.18053023889660835}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:03:42.779252+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026222.7792668, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.215108740244727, "elapsed_seconds": 4374.391464881017, "env_steps_per_second": 4114.858065289683, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.185374549731836, "eta_seconds": 15067.348379034613, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0001949902853466412, "chaser_grad_norm": 0.07723207026720047, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.71500015258789, "chaser_subspace_norm": 26.1453857421875, "collisions_per_episode": 1.2000000476837158, "entropy": 0.005384535761550069, "explorer_grad_norm": 0.9307530373334885, "explorer_new_fields": 97.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 94.9000015258789, "explorer_subspace_norm": 68.76335906982422, "final_distance": 2.279879331588745, "max_param_abs": 1.9441983699798584, "policy_loss": -0.00039095737622130144, "value_loss": 0.08582229167222977}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:11:42.621444+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026702.6214569, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3484726322866563, "elapsed_seconds": 4854.501476231962, "env_steps_per_second": 4119.887510163843, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.045417896859969, "eta_seconds": 14563.504428695887, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0003657927350999013, "chaser_grad_norm": 0.0461988365277648, "chaser_new_fields": 93.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.630000114440918, "chaser_subspace_norm": 27.59029197692871, "collisions_per_episode": 1.149999976158142, "entropy": 0.004460253054276109, "explorer_grad_norm": 0.8504244089126587, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.23750305175781, "explorer_subspace_norm": 70.28524017333984, "final_distance": 2.4047882556915283, "max_param_abs": 1.9868298768997192, "policy_loss": -0.0002500081267218235, "value_loss": 0.0880737155675888}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:42.731457+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027182.7314715, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.486390822277786, "elapsed_seconds": 5351.006960200029, "env_steps_per_second": 4111.375702486024, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.918666713277799, "eta_seconds": 14107.200167800076, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0011590251068150925, "chaser_grad_norm": 0.0902389707043767, "chaser_new_fields": 90.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.100000381469727, "chaser_subspace_norm": 28.05379295349121, "collisions_per_episode": 1.5, "entropy": 0.004823144758120179, "explorer_grad_norm": 1.4553553760051727, "explorer_new_fields": 97.1500015258789, "explorer_partner_vision": 0.0, "explorer_return": 94.9375, "explorer_subspace_norm": 76.13062286376953, "final_distance": 1.4426277875900269, "max_param_abs": 2.042477607727051, "policy_loss": -0.00040792157547286934, "value_loss": 0.15430257841944695}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:59.236940+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027679.2369516, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6218693919591736, "elapsed_seconds": 5838.729811053025, "env_steps_per_second": 4110.483063382507, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7843619145714045, "eta_seconds": 13623.702892457057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0003782547644554102, "chaser_grad_norm": 0.035297611029818654, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.714999198913574, "chaser_subspace_norm": 29.830278396606445, "collisions_per_episode": 0.875, "entropy": 0.004567723022773862, "explorer_grad_norm": 1.204826146364212, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 75.70590209960938, "final_distance": 2.658576250076294, "max_param_abs": 2.0844388008117676, "policy_loss": -0.0004188916831040501, "value_loss": 0.06732787378132343}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:36:06.959791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028166.959806, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7565214326358465, "elapsed_seconds": 6323.477157489047, "env_steps_per_second": 4111.66188355841, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6481598985513735, "eta_seconds": 13133.375634784945, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": -0.00011053696354215714, "chaser_grad_norm": 0.0255981357768178, "chaser_new_fields": 93.6500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.664999961853027, "chaser_subspace_norm": 32.38444137573242, "collisions_per_episode": 0.925000011920929, "entropy": 0.0029354990110732615, "explorer_grad_norm": 0.6075794547796249, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 81.52951049804688, "final_distance": 2.6078362464904785, "max_param_abs": 2.134159564971924, "policy_loss": -0.0002811870656742599, "value_loss": 0.06989477202296257}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:44:11.707136+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028651.7071483, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8909583165452932, "elapsed_seconds": 6807.4499395630555, "env_steps_per_second": 4113.140786724201, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5117797307269734, "eta_seconds": 12642.407030617103, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0009577972290002457, "chaser_grad_norm": 0.10972016118466854, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.680000305175781, "chaser_subspace_norm": 33.098140716552734, "collisions_per_episode": 0.6000000238418579, "entropy": 0.007064536679536104, "explorer_grad_norm": 1.555832326412201, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 81.40959167480469, "final_distance": 2.8271186351776123, "max_param_abs": 2.177057981491089, "policy_loss": -0.0009038806338788063, "value_loss": 0.0684131421148777}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:52:15.679922+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029135.6799397, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0263375222405497, "elapsed_seconds": 7294.815080065979, "env_steps_per_second": 4112.510004808602, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.377229203734249, "eta_seconds": 12158.025133443296, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0019205897785375958, "chaser_grad_norm": 0.07096149027347565, "chaser_new_fields": 93.5999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 35.65391540527344, "collisions_per_episode": 0.574999988079071, "entropy": 0.003614555753301829, "explorer_grad_norm": 0.6258119493722916, "explorer_new_fields": 97.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.98750305175781, "explorer_subspace_norm": 83.72328186035156, "final_distance": 2.50787615776062, "max_param_abs": 2.207871198654175, "policy_loss": -1.6525030186897993e-05, "value_loss": 0.05712071619927883}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:00:23.045059+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029623.0450718, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1620796764811128, "elapsed_seconds": 7783.486835332005, "env_steps_per_second": 4111.267954452066, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.243119514721669, "eta_seconds": 11675.230252998008, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0001679026009752116, "chaser_grad_norm": 0.04272673837840557, "chaser_new_fields": 94.17500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 37.833377838134766, "collisions_per_episode": 0.949999988079071, "entropy": 0.0038225125754252076, "explorer_grad_norm": 0.4116251841187477, "explorer_new_fields": 97.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.42500305175781, "explorer_subspace_norm": 86.74019622802734, "final_distance": 2.485421895980835, "max_param_abs": 2.244671106338501, "policy_loss": -0.00027566672084411437, "value_loss": 0.07080576568841934}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:08:31.716816+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030111.71683, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.297064289813053, "elapsed_seconds": 8269.43144332699, "env_steps_per_second": 4111.528130200084, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.107792862688248, "eta_seconds": 11188.054305677691, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0002199609705337835, "chaser_grad_norm": 0.052253819070756435, "chaser_new_fields": 94.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.789999961853027, "chaser_subspace_norm": 39.75495910644531, "collisions_per_episode": 0.7250000238418579, "entropy": 0.003014865389559418, "explorer_grad_norm": 0.5085247941315174, "explorer_new_fields": 97.75, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 86.49215698242188, "final_distance": 2.4540345668792725, "max_param_abs": 2.278742790222168, "policy_loss": -4.265964118577159e-05, "value_loss": 0.04733559302985668}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:16:37.661426+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030597.661439, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.431317870836388, "elapsed_seconds": 8752.744335010997, "env_steps_per_second": 4112.99572135335, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.971610731022252, "eta_seconds": 10697.798631680107, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 3.0737031003480464e-05, "chaser_grad_norm": 0.028669522609561682, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.8100004196167, "chaser_subspace_norm": 40.25957489013672, "collisions_per_episode": 0.925000011920929, "entropy": 0.0015017327095847577, "explorer_grad_norm": 0.6969778388738632, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 91.57691955566406, "final_distance": 2.3624792098999023, "max_param_abs": 2.311215877532959, "policy_loss": -9.36379457490144e-05, "value_loss": 0.07937017641961575}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:40.974316+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031080.9743292, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5684569116502662, "elapsed_seconds": 9246.444881940959, "env_steps_per_second": 4109.687613475858, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.838820797087137, "eta_seconds": 10219.754869513692, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00012275446308329663, "chaser_grad_norm": 0.0489763543009758, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.860000610351562, "chaser_subspace_norm": 41.91635513305664, "collisions_per_episode": 0.5250000357627869, "entropy": 0.003565317776519805, "explorer_grad_norm": 0.7209659665822983, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 93.65701293945312, "final_distance": 2.57779860496521, "max_param_abs": 2.3473610877990723, "policy_loss": -0.00027855207503257873, "value_loss": 0.04581528343260288}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:54.674863+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031574.6748776, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.705933880045292, "elapsed_seconds": 9741.361968163052, "env_steps_per_second": 4106.202000370065, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.705933880045292, "eta_seconds": 9741.361968163052, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00010105694170214896, "chaser_grad_norm": 0.060293670278042555, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.854999542236328, "chaser_subspace_norm": 43.29348373413086, "collisions_per_episode": 0.6500000357627869, "entropy": 0.003350229701027274, "explorer_grad_norm": 0.45895472168922424, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 96.93545532226562, "final_distance": 2.4522151947021484, "max_param_abs": 2.378573179244995, "policy_loss": -9.094262668796205e-05, "value_loss": 0.05636599659919739}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:41:09.591949+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032069.5919642, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8425155879594644, "elapsed_seconds": 10233.056116654072, "env_steps_per_second": 4104.3457126797075, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.571799817677611, "eta_seconds": 9258.4793436394, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0003398330166197727, "chaser_grad_norm": 0.30775879696011543, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.835000038146973, "chaser_subspace_norm": 44.78968811035156, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0016314693784806877, "explorer_grad_norm": 0.6310562491416931, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 95.13850402832031, "final_distance": 3.0631515979766846, "max_param_abs": 2.414945602416992, "policy_loss": -7.327461332096163e-05, "value_loss": 0.07577312365174294}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:49:21.286096+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032561.2861085, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.978737864720008, "elapsed_seconds": 10723.456312992028, "env_steps_per_second": 4103.154684063169, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4371491620436427, "eta_seconds": 8773.736983357114, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00015488800874823017, "chaser_grad_norm": 0.0528469355776906, "chaser_new_fields": 94.125, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 46.72232437133789, "collisions_per_episode": 0.625, "entropy": 0.0025998817873187363, "explorer_grad_norm": 0.33462097868323326, "explorer_new_fields": 97.25, "explorer_partner_vision": 0.0, "explorer_return": 95.5, "explorer_subspace_norm": 101.46304321289062, "final_distance": 2.472703695297241, "max_param_abs": 2.4533932209014893, "policy_loss": -0.00017467571358453782, "value_loss": 0.06614300236105919}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:31.686294+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033051.686308, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1163881493286074, "elapsed_seconds": 11218.997337582987, "env_steps_per_second": 4100.1881554871825, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.303417327764623, "eta_seconds": 8292.302379952644, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.00032714650077569907, "chaser_grad_norm": 0.05091250827535987, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 48.09226989746094, "collisions_per_episode": 0.75, "entropy": 0.004900307743810117, "explorer_grad_norm": 0.8155922293663025, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 101.57363891601562, "final_distance": 2.406571388244629, "max_param_abs": 2.483431816101074, "policy_loss": -0.0003047957538515611, "value_loss": 0.06800576485693455}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:47.227319+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033547.2273326, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2526695186227963, "elapsed_seconds": 11709.610267042066, "env_steps_per_second": 4099.197061673442, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1684463457485306, "eta_seconds": 7806.40684469471, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00046368235900579036, "chaser_grad_norm": 0.05172943649813533, "chaser_new_fields": 94.2750015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 47.75569152832031, "collisions_per_episode": 0.7250000238418579, "entropy": 0.004529269644990563, "explorer_grad_norm": 0.4677705317735672, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.61250305175781, "explorer_subspace_norm": 105.59456634521484, "final_distance": 2.6722252368927, "max_param_abs": 2.511936664581299, "policy_loss": -0.00021305633586621298, "value_loss": 0.04861252009868622}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:13:57.840248+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034037.840264, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3863353533425187, "elapsed_seconds": 12190.807272033067, "env_steps_per_second": 4101.451108550047, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0318012120055116, "eta_seconds": 7314.484363219841, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.002271625541464428, "chaser_grad_norm": 0.04877390363253653, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 49.18193435668945, "collisions_per_episode": 1.0500000715255737, "entropy": 0.004324603476561606, "explorer_grad_norm": 0.31020670384168625, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 105.13518524169922, "final_distance": 2.56154465675354, "max_param_abs": 2.539069652557373, "policy_loss": -0.0002574353291429077, "value_loss": 0.06831737235188484}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:21:59.037252+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034519.037263, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.521146916737489, "elapsed_seconds": 12676.12890025496, "env_steps_per_second": 4102.198739786726, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8960021859355713, "eta_seconds": 6825.6078693680565, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0013115831617422496, "chaser_grad_norm": 0.10855850018560886, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.824999809265137, "chaser_subspace_norm": 49.83440017700195, "collisions_per_episode": 0.75, "entropy": 0.003428350610192865, "explorer_grad_norm": 0.47709012031555176, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.61250305175781, "explorer_subspace_norm": 101.76946258544922, "final_distance": 2.5400571823120117, "max_param_abs": 2.5570921897888184, "policy_loss": -0.0007992418392097633, "value_loss": 0.07848214544355869}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:30:04.358883+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035004.3588982, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.656820060322223, "elapsed_seconds": 13164.552217160002, "env_steps_per_second": 4101.924555368542, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7606911401551444, "eta_seconds": 6338.48810455852, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": -1.1831437023102076e-05, "chaser_grad_norm": 0.11639629490673542, "chaser_new_fields": 93.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.730000495910645, "chaser_subspace_norm": 51.81495666503906, "collisions_per_episode": 1.100000023841858, "entropy": 0.004023518064059317, "explorer_grad_norm": 0.3094038739800453, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.5, "explorer_subspace_norm": 107.8450927734375, "final_distance": 2.8698978424072266, "max_param_abs": 2.5925300121307373, "policy_loss": -0.0001084711617735934, "value_loss": 0.07314801029860973}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:38:12.782197+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035492.7822077, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7905076778627698, "elapsed_seconds": 13645.827640305972, "env_steps_per_second": 4103.818506001909, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6245032905126155, "eta_seconds": 5848.211845845416, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0003166085974271038, "chaser_grad_norm": 0.039179712533950806, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.835000038146973, "chaser_subspace_norm": 51.99165725708008, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0023489276645705104, "explorer_grad_norm": 0.2554412707686424, "explorer_new_fields": 97.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.4625015258789, "explorer_subspace_norm": 107.26302337646484, "final_distance": 2.796588897705078, "max_param_abs": 2.617875814437866, "policy_loss": -8.236408825901265e-05, "value_loss": 0.07416762225329876}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:46:14.057623+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035974.0576382, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9245528739747373, "elapsed_seconds": 14128.390346309054, "env_steps_per_second": 4105.209339374751, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.488623503921452, "eta_seconds": 5359.044614117227, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -2.7392132638220934e-05, "chaser_grad_norm": 0.014969751238822937, "chaser_new_fields": 94.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.765000343322754, "chaser_subspace_norm": 51.79111862182617, "collisions_per_episode": 0.925000011920929, "entropy": 0.003710240009240806, "explorer_grad_norm": 0.351969912648201, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.5374984741211, "explorer_subspace_norm": 104.03763580322266, "final_distance": 3.044785737991333, "max_param_abs": 2.650078535079956, "policy_loss": -4.9626590752360755e-05, "value_loss": 0.07118609547615051}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:54:16.620326+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036456.6203384, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.058701530724454, "elapsed_seconds": 14611.325510608032, "env_steps_per_second": 4106.403622069684, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3529005102414844, "eta_seconds": 4870.441836869344, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00012317623491991947, "chaser_grad_norm": 0.019049633760005236, "chaser_new_fields": 94.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.8100004196167, "chaser_subspace_norm": 55.08219528198242, "collisions_per_episode": 0.8500000238418579, "entropy": 0.001874604175100103, "explorer_grad_norm": 0.28162549808621407, "explorer_new_fields": 97.75, "explorer_partner_vision": 0.0, "explorer_return": 95.9124984741211, "explorer_subspace_norm": 112.03358459472656, "final_distance": 2.484281539916992, "max_param_abs": 2.680609941482544, "policy_loss": -0.00021016992803346568, "value_loss": 0.061570482328534126}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:02:19.555491+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036939.5555081, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.19239851844138, "elapsed_seconds": 15092.63466638897, "env_steps_per_second": 4107.9640083035265, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2171479569668526, "eta_seconds": 4381.7326450806695, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.000361056688601466, "chaser_grad_norm": 0.06976673565804958, "chaser_new_fields": 94.25, "chaser_partner_vision": 0.0, "chaser_return": 8.8100004196167, "chaser_subspace_norm": 56.41948699951172, "collisions_per_episode": 0.800000011920929, "entropy": 0.0022809571819379926, "explorer_grad_norm": 0.3793414384126663, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.76250457763672, "explorer_subspace_norm": 108.27102661132812, "final_distance": 2.817970037460327, "max_param_abs": 2.699082374572754, "policy_loss": -0.0003219934841620642, "value_loss": 0.05619548819959164}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:10:20.864646+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037420.8646595, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.325781267537774, "elapsed_seconds": 15572.812563135987, "env_steps_per_second": 4109.726469802957, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0814453168844438, "eta_seconds": 3893.2031407839972, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": -1.4626533664685581e-05, "chaser_grad_norm": 0.04005991807207465, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.839999198913574, "chaser_subspace_norm": 55.52926254272461, "collisions_per_episode": 0.699999988079071, "entropy": 0.004258746048435569, "explorer_grad_norm": 0.3692924790084362, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.7249984741211, "explorer_subspace_norm": 113.34357452392578, "final_distance": 2.6441495418548584, "max_param_abs": 2.7303695678710938, "policy_loss": -0.0002719631420013968, "value_loss": 0.06901569478213787}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:18:21.042544+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037901.042558, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.460516325907794, "elapsed_seconds": 16057.85877326806, "env_steps_per_second": 4110.137032084996, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9461701297380171, "eta_seconds": 3406.2124670568614, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 7.246086871602608e-06, "chaser_grad_norm": 0.047289954498410225, "chaser_new_fields": 94.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.800000190734863, "chaser_subspace_norm": 55.27186584472656, "collisions_per_episode": 1.0750000476837158, "entropy": 0.001501561258919537, "explorer_grad_norm": 0.6360674202442169, "explorer_new_fields": 97.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.3499984741211, "explorer_subspace_norm": 114.8626937866211, "final_distance": 2.596870183944702, "max_param_abs": 2.7588918209075928, "policy_loss": -1.085281342239064e-06, "value_loss": 0.0664066169410944}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:26:26.088752+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038386.0887642, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.595091363174724, "elapsed_seconds": 16542.328907429008, "env_steps_per_second": 4110.666664925386, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8108984758543631, "eta_seconds": 2919.234513075707, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.00019201208183861063, "chaser_grad_norm": 0.025969069683924317, "chaser_new_fields": 94.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.894999504089355, "chaser_subspace_norm": 56.63450241088867, "collisions_per_episode": 0.6500000357627869, "entropy": 0.002264530921820551, "explorer_grad_norm": 0.2563510462641716, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 111.8426513671875, "final_distance": 3.092700242996216, "max_param_abs": 2.777531623840332, "policy_loss": -0.0002753162422148847, "value_loss": 0.04043085779994726}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:30.558888+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038870.5589042, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.733458697345825, "elapsed_seconds": 17040.451310444972, "env_steps_per_second": 4107.8724221989, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6762083853351178, "eta_seconds": 2434.3501872064244, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0005584626693679873, "chaser_grad_norm": 0.026837595039978623, "chaser_new_fields": 94.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 57.78645706176758, "collisions_per_episode": 0.699999988079071, "entropy": 0.002988118852954358, "explorer_grad_norm": 0.2947659231722355, "explorer_new_fields": 97.3499984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.61250305175781, "explorer_subspace_norm": 115.66912078857422, "final_distance": 2.6776599884033203, "max_param_abs": 2.8079304695129395, "policy_loss": -0.0002786490936999009, "value_loss": 0.07245573587715626}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:48.681291+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039368.6813056, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.872163247556956, "elapsed_seconds": 17539.78769120504, "env_steps_per_second": 4104.952766110327, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5413514719507729, "eta_seconds": 1948.8652990227827, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 3.373953766028226e-05, "chaser_grad_norm": 0.05039068963378668, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.880000114440918, "chaser_subspace_norm": 57.42374038696289, "collisions_per_episode": 0.800000011920929, "entropy": 0.0018022336007561535, "explorer_grad_norm": 0.556768074631691, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 119.5387954711914, "final_distance": 2.658203601837158, "max_param_abs": 2.82814884185791, "policy_loss": -6.0003283302467736e-05, "value_loss": 0.07089593634009361}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:51:08.017674+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039868.017689, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.007059313671942, "elapsed_seconds": 18025.41352921899, "env_steps_per_second": 4105.314969892194, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40597778218961683, "eta_seconds": 1461.5200158826206, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0038805218834860877, "chaser_grad_norm": 0.22556009981781244, "chaser_new_fields": 93.95000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.725000381469727, "chaser_subspace_norm": 60.12946319580078, "collisions_per_episode": 0.8500000238418579, "entropy": 0.0030050898785702884, "explorer_grad_norm": 0.32349031046032906, "explorer_new_fields": 97.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 123.29110717773438, "final_distance": 2.3544228076934814, "max_param_abs": 2.841538667678833, "policy_loss": -0.0010514583834435598, "value_loss": 0.08223988115787506}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:59:13.643508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040353.6435208, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.142374565717215, "elapsed_seconds": 18512.548436581972, "env_steps_per_second": 4105.32349235177, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27065129293248497, "eta_seconds": 974.3446545569458, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -0.00014539641226721245, "chaser_grad_norm": 0.02960295882076025, "chaser_new_fields": 94.25, "chaser_partner_vision": 0.0, "chaser_return": 8.805000305175781, "chaser_subspace_norm": 59.347694396972656, "collisions_per_episode": 0.675000011920929, "entropy": 0.0016602027753833681, "explorer_grad_norm": 0.72232785820961, "explorer_new_fields": 97.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 118.7383041381836, "final_distance": 2.3005945682525635, "max_param_abs": 2.8610944747924805, "policy_loss": -0.00010950184566915766, "value_loss": 0.0588388666510582}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:07:20.778418+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040840.7784326, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.277653096073061, "elapsed_seconds": 18999.55114586302, "env_steps_per_second": 4105.360142520198, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13532443836084773, "eta_seconds": 487.1679780990518, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0001011030007935787, "chaser_grad_norm": 0.03486114228144288, "chaser_new_fields": 94.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.845000267028809, "chaser_subspace_norm": 61.14625930786133, "collisions_per_episode": 0.699999988079071, "entropy": 0.0014901134418323636, "explorer_grad_norm": 0.4387537054717541, "explorer_new_fields": 97.5, "explorer_partner_vision": 0.0, "explorer_return": 95.6500015258789, "explorer_subspace_norm": 121.42359161376953, "final_distance": 2.5939605236053467, "max_param_abs": 2.8831002712249756, "policy_loss": -0.00017680002559572472, "value_loss": 0.055940572172403336}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:15:27.781125+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041327.7811368, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.408693705739441, "elapsed_seconds": 19471.297340661986, "env_steps_per_second": 4108.611696506513, "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-transformer/non_social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:19.527319+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041799.5273309, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.408702007005275, "elapsed_seconds": 19471.327225218993, "env_steps_per_second": 4108.605390616882, "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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:19.557206+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041799.557218, "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..15e84b164692a90badeabe0daa31f557d31d6430 --- /dev/null +++ b/non_social/seed_0008/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.408702007005275, + "elapsed_seconds": 19471.327225218993, + "env_steps_per_second": 4108.605390616882, + "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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 6.500302986564566e-05, + "chaser_grad_norm": 0.026715639047324657, + "chaser_new_fields": 94.5, + "chaser_partner_vision": 0.0, + "chaser_return": 8.855000495910645, + "chaser_subspace_norm": 64.1446304321289, + "collisions_per_episode": 0.7250000238418579, + "entropy": 0.0010687967587728053, + "explorer_grad_norm": 0.17880532704293728, + "explorer_new_fields": 97.625, + "explorer_partner_vision": 0.0, + "explorer_return": 95.875, + "explorer_subspace_norm": 122.6617660522461, + "final_distance": 2.5540859699249268, + "max_param_abs": 2.898890733718872, + "policy_loss": -0.00012717402295669977, + "value_loss": 0.05584537237882614 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:19.557206+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041799.557218, + "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..0288af5dd39b9119815c8dad9025a5d6327cae69 --- /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 sha256:6c3ce75ac26f6605968ebb1613d72e3f3e59ac898be34d14b34a59812bd25a49 +size 29760344 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..469f1fef28d107755b26261b38c6887af7e31e4b --- /dev/null +++ b/non_social/seed_0009/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/non_social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..65f11bc4127297dc4eb2082778854183bb9a1f0a --- /dev/null +++ b/non_social/seed_0009/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006997621372122214, "elapsed_seconds": 2.519143693963997, "env_steps_per_second": 1587.8411420452965, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.994542982107216, "eta_seconds": 50380.35473558598, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.006501829170386664, "chaser_grad_norm": 8.34403159469366, "chaser_new_fields": 34.07500076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.2250003814697266, "chaser_subspace_norm": 0.03185378760099411, "collisions_per_episode": 1.9500000476837158, "entropy": 2.75326406955719, "explorer_grad_norm": 6.38175904750824, "explorer_new_fields": 36.32500076293945, "explorer_partner_vision": 0.0, "explorer_return": 3.924999952316284, "explorer_subspace_norm": 0.0466776117682457, "final_distance": 5.312294960021973, "max_param_abs": 1.0012015104293823, "policy_loss": -0.012270373390066425, "value_loss": 7.722126603126526}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:00:02.716259+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022402.7162719, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13650578946612466, "elapsed_seconds": 491.42084207804874, "env_steps_per_second": 4069.831453510787, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.323725789178861, "eta_seconds": 19165.4128410439, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005085971632653674, "chaser_grad_norm": 0.45659415796399117, "chaser_new_fields": 90.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.030000686645508, "chaser_subspace_norm": 13.179183959960938, "collisions_per_episode": 2.1000001430511475, "entropy": 0.12038099579513073, "explorer_grad_norm": 7.830848693847656, "explorer_new_fields": 87.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 81.4000015258789, "explorer_subspace_norm": 14.228251457214355, "final_distance": 3.3970839977264404, "max_param_abs": 1.2857152223587036, "policy_loss": -0.0020245612623419618, "value_loss": 0.6201536804437637}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:08:11.617958+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022891.617971, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27201322878109446, "elapsed_seconds": 979.2476236119401, "env_steps_per_second": 4084.768656620335, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.168251346840795, "eta_seconds": 18605.70484862686, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0010975377837052491, "chaser_grad_norm": 0.13131846487522125, "chaser_new_fields": 92.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.380000114440918, "chaser_subspace_norm": 19.698219299316406, "collisions_per_episode": 0.9000000357627869, "entropy": 0.02975855302065611, "explorer_grad_norm": 3.179869621992111, "explorer_new_fields": 93.70000457763672, "explorer_partner_vision": 0.0, "explorer_return": 89.95000457763672, "explorer_subspace_norm": 28.595848083496094, "final_distance": 2.9191970825195312, "max_param_abs": 1.4518451690673828, "policy_loss": -0.0011499320692394122, "value_loss": 0.2044721283018589}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:16:19.444737+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023379.4447527, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4071205642547122, "elapsed_seconds": 1465.634031316964, "env_steps_per_second": 4093.7914048083508, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.021153625808116, "eta_seconds": 18076.15305290922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0030521016570295156, "chaser_grad_norm": 0.24272240698337555, "chaser_new_fields": 92.1500015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.375, "chaser_subspace_norm": 22.737030029296875, "collisions_per_episode": 1.125, "entropy": 0.017216809559613466, "explorer_grad_norm": 2.593182533979416, "explorer_new_fields": 94.67500305175781, "explorer_partner_vision": 0.0, "explorer_return": 91.3375015258789, "explorer_subspace_norm": 38.73212814331055, "final_distance": 3.1398448944091797, "max_param_abs": 1.590619444847107, "policy_loss": -0.0025943185216874554, "value_loss": 0.24982118979096413}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:24:25.831147+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023865.8311594, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5425571607177779, "elapsed_seconds": 1953.2057785840007, "env_steps_per_second": 4095.830602037074, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.883014446460002, "eta_seconds": 17578.852007256006, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.001681200492763768, "chaser_grad_norm": 0.1912822090089321, "chaser_new_fields": 93.25, "chaser_partner_vision": 0.0, "chaser_return": 8.620000839233398, "chaser_subspace_norm": 27.78968620300293, "collisions_per_episode": 0.3499999940395355, "entropy": 0.007195838843472302, "explorer_grad_norm": 1.0246603786945343, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.6875, "explorer_subspace_norm": 49.048152923583984, "final_distance": 3.0071983337402344, "max_param_abs": 1.6958986520767212, "policy_loss": -0.0015016091470698711, "value_loss": 0.0831474307924509}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:32:33.402894+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024353.4029071, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6775989613861182, "elapsed_seconds": 2439.3562609900255, "env_steps_per_second": 4099.442201173783, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.743192729702828, "eta_seconds": 17075.49382693018, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.00016674404568872703, "chaser_grad_norm": 0.10249152779579163, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.635001182556152, "chaser_subspace_norm": 27.66214942932129, "collisions_per_episode": 0.4749999940395355, "entropy": 0.009833229705691338, "explorer_grad_norm": 1.782571405172348, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 56.74618148803711, "final_distance": 2.726036310195923, "max_param_abs": 1.7895199060440063, "policy_loss": -0.000762642422230897, "value_loss": 0.09628107771277428}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:40:39.553374+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024839.5533862, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8108764808530557, "elapsed_seconds": 2919.1553310710005, "env_steps_per_second": 4110.77816664088, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.5949667248339825, "eta_seconds": 16541.880209402338, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0008238025012277106, "chaser_grad_norm": 0.11354982107877731, "chaser_new_fields": 93.5, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 29.639259338378906, "collisions_per_episode": 1.0500000715255737, "entropy": 0.011348627973347902, "explorer_grad_norm": 1.2415755465626717, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.4124984741211, "explorer_subspace_norm": 60.66948699951172, "final_distance": 3.122871160507202, "max_param_abs": 1.8411896228790283, "policy_loss": -0.0024167038416607056, "value_loss": 0.221784807741642}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:48:39.352446+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025319.3524613, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9464378867563856, "elapsed_seconds": 3407.1763923229882, "env_steps_per_second": 4108.9742320194055, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.46177860899439, "eta_seconds": 16062.402992379803, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004445954239992833, "chaser_grad_norm": 0.05461905151605606, "chaser_new_fields": 92.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.510001182556152, "chaser_subspace_norm": 32.47053527832031, "collisions_per_episode": 1.225000023841858, "entropy": 0.009320953860878944, "explorer_grad_norm": 1.5863610953092575, "explorer_new_fields": 93.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.625, "explorer_subspace_norm": 67.61103057861328, "final_distance": 3.2263119220733643, "max_param_abs": 1.8798553943634033, "policy_loss": -0.0030323344012685993, "value_loss": 0.23431266099214554}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:56:47.373506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025807.3735173, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0809030087074887, "elapsed_seconds": 3891.250831346959, "env_steps_per_second": 4111.7883923359395, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.323612034829955, "eta_seconds": 15565.003325387836, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00040988103761113726, "chaser_grad_norm": 0.05791908595710993, "chaser_new_fields": 93.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.600000381469727, "chaser_subspace_norm": 34.47807312011719, "collisions_per_episode": 0.800000011920929, "entropy": 0.010967804817482829, "explorer_grad_norm": 1.3904897719621658, "explorer_new_fields": 94.75, "explorer_partner_vision": 0.0, "explorer_return": 91.63750457763672, "explorer_subspace_norm": 70.48702239990234, "final_distance": 3.2297515869140625, "max_param_abs": 1.9223394393920898, "policy_loss": -0.0006720104694837303, "value_loss": 0.15366922691464424}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:04:51.447945+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026291.4479597, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2178072636586148, "elapsed_seconds": 4384.106149171013, "env_steps_per_second": 4105.74000435724, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.194669463713007, "eta_seconds": 15100.810069366826, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0001055804356789869, "chaser_grad_norm": 0.03483629459515214, "chaser_new_fields": 93.92500305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.720000267028809, "chaser_subspace_norm": 35.2333869934082, "collisions_per_episode": 0.824999988079071, "entropy": 0.005636593094095588, "explorer_grad_norm": 0.7687147259712219, "explorer_new_fields": 95.0, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 76.42829132080078, "final_distance": 3.158228635787964, "max_param_abs": 1.953141212463379, "policy_loss": -0.00037110067933099344, "value_loss": 0.09170574322342873}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:13:04.303264+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026784.3032777, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.352685223195262, "elapsed_seconds": 4869.666803502943, "env_steps_per_second": 4107.057178042081, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.0580556695857855, "eta_seconds": 14609.000410508828, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0005116107642564316, "chaser_grad_norm": 0.051148759201169014, "chaser_new_fields": 93.32500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.610000610351562, "chaser_subspace_norm": 39.25312805175781, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0068903082283213735, "explorer_grad_norm": 0.3995022401213646, "explorer_new_fields": 94.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 91.75, "explorer_subspace_norm": 81.02411651611328, "final_distance": 3.301560640335083, "max_param_abs": 1.9766359329223633, "policy_loss": -0.0006573367682420894, "value_loss": 0.09519045613706112}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:21:09.863918+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027269.863934, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4864643141763776, "elapsed_seconds": 5351.2715310349595, "env_steps_per_second": 4111.172433020812, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.918860464646814, "eta_seconds": 14107.89767272853, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0040434074320988556, "chaser_grad_norm": 0.11104412470012903, "chaser_new_fields": 93.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.625, "chaser_subspace_norm": 40.41009521484375, "collisions_per_episode": 1.0, "entropy": 0.008885535644367337, "explorer_grad_norm": 0.9410631060600281, "explorer_new_fields": 94.57500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.3375015258789, "explorer_subspace_norm": 82.81352996826172, "final_distance": 3.8293023109436035, "max_param_abs": 2.014787197113037, "policy_loss": -0.0018034361061154414, "value_loss": 0.14407268166542053}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:29:11.468644+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027751.4686573, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6231152229972132, "elapsed_seconds": 5843.2148027899675, "env_steps_per_second": 4107.328039445117, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.787268853660164, "eta_seconds": 13634.16787317659, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0005749695391921716, "chaser_grad_norm": 0.05962391383945942, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.655000686645508, "chaser_subspace_norm": 41.88115692138672, "collisions_per_episode": 0.6000000238418579, "entropy": 0.0036552156088873744, "explorer_grad_norm": 0.3345196954905987, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 87.39372253417969, "final_distance": 3.5411343574523926, "max_param_abs": 2.037968635559082, "policy_loss": -8.502483849592846e-05, "value_loss": 0.07618293911218643}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:37:23.411919+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028243.4119334, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.759172244602782, "elapsed_seconds": 6333.020080570015, "env_steps_per_second": 4105.466218205931, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6536654310980854, "eta_seconds": 13153.195551953107, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0007038068265348141, "chaser_grad_norm": 0.13297725096344948, "chaser_new_fields": 93.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.645000457763672, "chaser_subspace_norm": 45.23651885986328, "collisions_per_episode": 0.8500000238418579, "entropy": 0.0036652859998866916, "explorer_grad_norm": 0.4156603515148163, "explorer_new_fields": 95.0250015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 88.31893157958984, "final_distance": 3.3274269104003906, "max_param_abs": 2.0707902908325195, "policy_loss": -0.0007617957714263213, "value_loss": 0.12042364291846752}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:45:33.217194+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028733.2172055, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.894276243340832, "elapsed_seconds": 6819.394476026995, "env_steps_per_second": 4105.936399255334, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.517941594775831, "eta_seconds": 12664.589741192991, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.000629503821953814, "chaser_grad_norm": 0.06746643874794245, "chaser_new_fields": 94.9749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.964999198913574, "chaser_subspace_norm": 43.46806716918945, "collisions_per_episode": 0.42500001192092896, "entropy": 0.005476371035911143, "explorer_grad_norm": 0.837139442563057, "explorer_new_fields": 94.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 89.13663482666016, "final_distance": 3.044492721557617, "max_param_abs": 2.0918331146240234, "policy_loss": -0.0008937543926297309, "value_loss": 0.05374081805348396}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:53:39.591590+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029219.5916042, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.029450304734734, "elapsed_seconds": 7306.021097045043, "env_steps_per_second": 4106.202213422796, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.38241717455789, "eta_seconds": 12176.701828408404, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0003893071452768515, "chaser_grad_norm": 0.036772985477000475, "chaser_new_fields": 95.30000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.029999732971191, "chaser_subspace_norm": 49.0320930480957, "collisions_per_episode": 0.550000011920929, "entropy": 0.0043357707327231765, "explorer_grad_norm": 0.3147982731461525, "explorer_new_fields": 95.32500457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.76250457763672, "explorer_subspace_norm": 92.92819213867188, "final_distance": 3.020263195037842, "max_param_abs": 2.10390567779541, "policy_loss": -0.00043892076029550253, "value_loss": 0.06615459732711315}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:01:46.218210+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029706.2182224, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.162906503881386, "elapsed_seconds": 7786.46341397299, "env_steps_per_second": 4109.696315091554, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.244359755822079, "eta_seconds": 11679.695120959484, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.002330733255991535, "chaser_grad_norm": 0.10410290025174618, "chaser_new_fields": 94.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.910000801086426, "chaser_subspace_norm": 48.768836975097656, "collisions_per_episode": 0.8500000238418579, "entropy": 0.009912226349115372, "explorer_grad_norm": 0.7728115171194077, "explorer_new_fields": 93.9749984741211, "explorer_partner_vision": 0.0, "explorer_return": 90.4749984741211, "explorer_subspace_norm": 93.35250854492188, "final_distance": 3.2252578735351562, "max_param_abs": 2.1016077995300293, "policy_loss": -0.003898917519492784, "value_loss": 0.27449797093868256}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:09:46.660528+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030186.660542, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.2988223328260937, "elapsed_seconds": 8275.760398173938, "env_steps_per_second": 4108.3838057348985, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.110171391470597, "eta_seconds": 11196.61700929415, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0015037483692658116, "chaser_grad_norm": 0.07312043104320765, "chaser_new_fields": 94.8499984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.930000305175781, "chaser_subspace_norm": 52.66928482055664, "collisions_per_episode": 0.699999988079071, "entropy": 0.00904917810112238, "explorer_grad_norm": 1.572329342365265, "explorer_new_fields": 94.7249984741211, "explorer_partner_vision": 0.0, "explorer_return": 91.7125015258789, "explorer_subspace_norm": 92.63137817382812, "final_distance": 3.323681592941284, "max_param_abs": 2.12587308883667, "policy_loss": -0.0017253117980402521, "value_loss": 0.20381438359618187}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:17:55.957511+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030675.9575233, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4325308703527684, "elapsed_seconds": 8757.111133269966, "env_steps_per_second": 4110.944745605547, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.973093285986717, "eta_seconds": 10703.13582955218, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0005336430528459779, "chaser_grad_norm": 0.03605387918651104, "chaser_new_fields": 94.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.885000228881836, "chaser_subspace_norm": 57.63795471191406, "collisions_per_episode": 0.675000011920929, "entropy": 0.005303243990056217, "explorer_grad_norm": 0.273235734552145, "explorer_new_fields": 95.25, "explorer_partner_vision": 0.0, "explorer_return": 92.57500457763672, "explorer_subspace_norm": 97.46886444091797, "final_distance": 2.8869431018829346, "max_param_abs": 2.140315294265747, "policy_loss": -0.00028817273472725446, "value_loss": 0.10296069458127022}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:25:57.308248+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031157.3082623, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.565980261175007, "elapsed_seconds": 9237.528940230026, "env_steps_per_second": 4113.654230029806, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8360834465618505, "eta_seconds": 10209.900407622661, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0003900933318394806, "chaser_grad_norm": 0.034444418735802174, "chaser_new_fields": 94.4749984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 57.71100616455078, "collisions_per_episode": 0.699999988079071, "entropy": 0.006893767626024783, "explorer_grad_norm": 0.4262780025601387, "explorer_new_fields": 95.125, "explorer_partner_vision": 0.0, "explorer_return": 92.3125, "explorer_subspace_norm": 100.76033782958984, "final_distance": 3.065006971359253, "max_param_abs": 2.1518313884735107, "policy_loss": -0.000673958832983601, "value_loss": 0.15167176350951195}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:33:57.726053+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031637.726066, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7007485915238956, "elapsed_seconds": 9722.694929486024, "env_steps_per_second": 4114.085682015176, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7007485915238956, "eta_seconds": 9722.694929486024, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0011435146810068186, "chaser_grad_norm": 0.03576818015426397, "chaser_new_fields": 95.0250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.964999198913574, "chaser_subspace_norm": 58.145896911621094, "collisions_per_episode": 0.75, "entropy": 0.006610351498238742, "explorer_grad_norm": 0.3946061059832573, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.125, "explorer_subspace_norm": 101.43132781982422, "final_distance": 3.454437255859375, "max_param_abs": 2.1779396533966064, "policy_loss": -0.0005541431005102737, "value_loss": 0.10217516124248505}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:42:02.892044+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032122.8920596, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8340633647041655, "elapsed_seconds": 10202.628112934995, "env_steps_per_second": 4116.586386869475, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.564152568065673, "eta_seconds": 9230.949245036423, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0009154198487488774, "chaser_grad_norm": 0.03069269424304366, "chaser_new_fields": 95.20000457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.99000072479248, "chaser_subspace_norm": 59.35086441040039, "collisions_per_episode": 0.699999988079071, "entropy": 0.004197821719571948, "explorer_grad_norm": 0.40724140778183937, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 92.9124984741211, "explorer_subspace_norm": 106.12228393554688, "final_distance": 3.2408998012542725, "max_param_abs": 2.199213981628418, "policy_loss": -0.000907265011917957, "value_loss": 0.051750375889241695}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:50:02.825226+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032602.8252385, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9705745488941706, "elapsed_seconds": 10694.068376019015, "env_steps_per_second": 4114.430397571433, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4304700854588672, "eta_seconds": 8749.692307651922, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0002962914765736735, "chaser_grad_norm": 0.020425160648301244, "chaser_new_fields": 95.2249984741211, "chaser_partner_vision": 0.0, "chaser_return": 9.0, "chaser_subspace_norm": 63.95237350463867, "collisions_per_episode": 0.625, "entropy": 0.002526312891859561, "explorer_grad_norm": 0.5745395496487617, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 108.66472625732422, "final_distance": 3.6302497386932373, "max_param_abs": 2.215160369873047, "policy_loss": -0.0001902093905082891, "value_loss": 0.08152232319116592}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:58:14.265490+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033094.2655058, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1068204238249906, "elapsed_seconds": 11184.553525769967, "env_steps_per_second": 4112.815043891818, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.2963455306532543, "eta_seconds": 8266.843910351716, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0004841115155479678, "chaser_grad_norm": 0.030746271600946784, "chaser_new_fields": 95.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 9.055000305175781, "chaser_subspace_norm": 65.43505096435547, "collisions_per_episode": 0.574999988079071, "entropy": 0.005808753194287419, "explorer_grad_norm": 0.28011006489396095, "explorer_new_fields": 95.2750015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.6500015258789, "explorer_subspace_norm": 108.83171844482422, "final_distance": 4.134437084197998, "max_param_abs": 2.2292299270629883, "policy_loss": -0.0005104938026079253, "value_loss": 0.10148402489721775}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:06:24.750639+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033584.7506511, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2401549243600067, "elapsed_seconds": 11664.557727696025, "env_steps_per_second": 4115.029572534074, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1601032829066713, "eta_seconds": 7776.3718184640165, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 3.465173754678419e-05, "chaser_grad_norm": 0.019946390064433217, "chaser_new_fields": 95.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 9.045001029968262, "chaser_subspace_norm": 65.08322143554688, "collisions_per_episode": 0.625, "entropy": 0.0025112821022048593, "explorer_grad_norm": 0.5748406052589417, "explorer_new_fields": 95.375, "explorer_partner_vision": 0.0, "explorer_return": 92.61250305175781, "explorer_subspace_norm": 109.34302520751953, "final_distance": 4.334353923797607, "max_param_abs": 2.2406883239746094, "policy_loss": -7.82804499976919e-05, "value_loss": 0.06383916083723307}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:14:24.754842+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034064.7548559, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.375532995280292, "elapsed_seconds": 12151.918783009052, "env_steps_per_second": 4114.57654489187, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.025319797168175, "eta_seconds": 7291.15126980543, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00103564713427734, "chaser_grad_norm": 0.07754432689398527, "chaser_new_fields": 95.375, "chaser_partner_vision": 0.0, "chaser_return": 9.030000686645508, "chaser_subspace_norm": 71.1597900390625, "collisions_per_episode": 0.7250000238418579, "entropy": 0.002828277414664626, "explorer_grad_norm": 0.4504784904420376, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.98750305175781, "explorer_subspace_norm": 114.89822387695312, "final_distance": 4.206541538238525, "max_param_abs": 2.2617409229278564, "policy_loss": -0.0004102075305993935, "value_loss": 0.07798822969198227}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:22:32.115899+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034552.1159117, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5110608257927622, "elapsed_seconds": 12639.818972853944, "env_steps_per_second": 4113.9829701420895, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8905712138884103, "eta_seconds": 6806.056369998277, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0014179340557183007, "chaser_grad_norm": 0.04576253332197666, "chaser_new_fields": 95.0999984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.970000267028809, "chaser_subspace_norm": 71.85916137695312, "collisions_per_episode": 0.675000011920929, "entropy": 0.004850554862059653, "explorer_grad_norm": 0.25398726388812065, "explorer_new_fields": 95.75, "explorer_partner_vision": 0.0, "explorer_return": 93.32500457763672, "explorer_subspace_norm": 115.05350494384766, "final_distance": 4.191352367401123, "max_param_abs": 2.275130271911621, "policy_loss": -0.000501708065391071, "value_loss": 0.061753932386636734}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:30:40.016088+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035040.016103, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6478938717933165, "elapsed_seconds": 13132.41793845594, "env_steps_per_second": 4111.961731119648, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7563933456782632, "eta_seconds": 6323.016044441748, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0008805659902768603, "chaser_grad_norm": 0.09235539007931948, "chaser_new_fields": 95.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.039999961853027, "chaser_subspace_norm": 79.50959777832031, "collisions_per_episode": 0.42500001192092896, "entropy": 0.0031440184102393687, "explorer_grad_norm": 0.453799020498991, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 120.18220520019531, "final_distance": 4.107096195220947, "max_param_abs": 2.287307024002075, "policy_loss": -0.00042727127450348235, "value_loss": 0.0608030017465353}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:38:52.615054+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035532.6150682, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7857993634710954, "elapsed_seconds": 13628.877708495944, "env_steps_per_second": 4108.922333721641, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6224854414876124, "eta_seconds": 5840.9475893554045, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": -2.5781603289004584e-05, "chaser_grad_norm": 0.050726886838674545, "chaser_new_fields": 95.55000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.080000877380371, "chaser_subspace_norm": 78.46027374267578, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0012880461290478706, "explorer_grad_norm": 0.1940711885690689, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.25, "explorer_subspace_norm": 117.79642486572266, "final_distance": 3.7183284759521484, "max_param_abs": 2.297485589981079, "policy_loss": -3.759408264891739e-05, "value_loss": 0.05101415328681469}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:47:09.074824+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036029.0748372, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.919367772837221, "elapsed_seconds": 14109.723982213996, "env_steps_per_second": 4110.6402983582, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.486656741421015, "eta_seconds": 5351.964269115654, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00037415088831354443, "chaser_grad_norm": 0.2084239087998867, "chaser_new_fields": 95.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 9.045000076293945, "chaser_subspace_norm": 79.32145690917969, "collisions_per_episode": 0.4749999940395355, "entropy": 0.002682093298062682, "explorer_grad_norm": 0.29349998012185097, "explorer_new_fields": 95.6500015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.32500457763672, "explorer_subspace_norm": 122.88455963134766, "final_distance": 4.297656536102295, "max_param_abs": 2.3066563606262207, "policy_loss": -0.00035055562213415215, "value_loss": 0.051116110756993294}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:55:09.921097+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036509.9211085, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.055394076383899, "elapsed_seconds": 14599.418674982036, "env_steps_per_second": 4109.752678222568, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3517980254612998, "eta_seconds": 4866.472891660679, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0008715333203519471, "chaser_grad_norm": 0.06456276262179017, "chaser_new_fields": 95.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 9.080000877380371, "chaser_subspace_norm": 79.70638275146484, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0038003838853910565, "explorer_grad_norm": 0.31966231390833855, "explorer_new_fields": 95.4000015258789, "explorer_partner_vision": 0.0, "explorer_return": 92.875, "explorer_subspace_norm": 124.3276138305664, "final_distance": 3.7052230834960938, "max_param_abs": 2.3284823894500732, "policy_loss": -0.0014383688620327462, "value_loss": 0.06927209161221981}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:03:19.615790+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036999.615806, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.188820491956948, "elapsed_seconds": 15079.753771045012, "env_steps_per_second": 4111.4729684146205, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2161091750842752, "eta_seconds": 4377.993030303391, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0003140873018335818, "chaser_grad_norm": 0.030534851364791393, "chaser_new_fields": 95.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 9.095000267028809, "chaser_subspace_norm": 83.11844635009766, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0025616076891310513, "explorer_grad_norm": 0.31148969382047653, "explorer_new_fields": 95.17500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.5, "explorer_subspace_norm": 126.38959503173828, "final_distance": 4.310662746429443, "max_param_abs": 2.3365540504455566, "policy_loss": -0.00043565417779944937, "value_loss": 0.0858713686466217}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:11:19.950884+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037479.9508967, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.322445538695288, "elapsed_seconds": 15560.803939303034, "env_steps_per_second": 4112.898038535825, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.080611384673822, "eta_seconds": 3890.200984825759, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0034916070632977636, "chaser_grad_norm": 0.13872127700597048, "chaser_new_fields": 95.45000457763672, "chaser_partner_vision": 0.0, "chaser_return": 9.0600004196167, "chaser_subspace_norm": 86.72901916503906, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0028467164374887943, "explorer_grad_norm": 0.47973188012838364, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.80000305175781, "explorer_subspace_norm": 122.14115142822266, "final_distance": 4.2479166984558105, "max_param_abs": 2.3535091876983643, "policy_loss": -0.001579295195638153, "value_loss": 0.07589246705174446}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:19:21.001053+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037961.0010679, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.456240056085549, "elapsed_seconds": 16042.464201907977, "env_steps_per_second": 4114.081176640583, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9452630421999649, "eta_seconds": 3402.9469519198738, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0005584280076086356, "chaser_grad_norm": 0.09734787698835135, "chaser_new_fields": 95.42500305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.045000076293945, "chaser_subspace_norm": 83.404541015625, "collisions_per_episode": 0.824999988079071, "entropy": 0.002018804050749168, "explorer_grad_norm": 0.2797966618090868, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.23750305175781, "explorer_subspace_norm": 123.76571655273438, "final_distance": 4.201670169830322, "max_param_abs": 2.3649237155914307, "policy_loss": -0.0004505110061727713, "value_loss": 0.10733920894563198}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:27:22.661316+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038442.6613297, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.58971992157778, "elapsed_seconds": 16522.991717680008, "env_steps_per_second": 4115.477460854642, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8099505743960788, "eta_seconds": 2915.8220678258836, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 4.2007002676959504e-05, "chaser_grad_norm": 0.061925873160362244, "chaser_new_fields": 95.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 9.045000076293945, "chaser_subspace_norm": 84.53114318847656, "collisions_per_episode": 0.574999988079071, "entropy": 0.0017206192715093493, "explorer_grad_norm": 0.34876298904418945, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 126.01876068115234, "final_distance": 4.062276363372803, "max_param_abs": 2.3885884284973145, "policy_loss": -9.286738150660057e-05, "value_loss": 0.05439954437315464}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:35:23.188832+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038923.188846, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.724103955824168, "elapsed_seconds": 17006.774240967003, "env_steps_per_second": 4116.006892793316, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6748719936891667, "eta_seconds": 2429.5391772810003, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 5.352570752359398e-05, "chaser_grad_norm": 0.03992908028885722, "chaser_new_fields": 95.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 9.144999504089355, "chaser_subspace_norm": 86.08087921142578, "collisions_per_episode": 0.6500000357627869, "entropy": 0.0016274515073746443, "explorer_grad_norm": 0.39181549847126007, "explorer_new_fields": 95.0999984741211, "explorer_partner_vision": 0.0, "explorer_return": 92.42500305175781, "explorer_subspace_norm": 128.2463836669922, "final_distance": 4.284608364105225, "max_param_abs": 2.4417428970336914, "policy_loss": -0.00025753481709145376, "value_loss": 0.07663443684577942}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:43:26.971355+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039406.971366, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.857896780363614, "elapsed_seconds": 17488.42840930901, "env_steps_per_second": 4117.008018952391, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5397663089292905, "eta_seconds": 1943.1587121454456, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 3.83070747123071e-05, "chaser_grad_norm": 0.02687141438946128, "chaser_new_fields": 95.70000457763672, "chaser_partner_vision": 0.0, "chaser_return": 9.105000495910645, "chaser_subspace_norm": 85.44757843017578, "collisions_per_episode": 0.574999988079071, "entropy": 0.0026248095091432333, "explorer_grad_norm": 0.193992767482996, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.8375015258789, "explorer_subspace_norm": 130.38284301757812, "final_distance": 4.5929999351501465, "max_param_abs": 2.4926345348358154, "policy_loss": -0.0001301567051268382, "value_loss": 0.10110200569033623}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:51:28.625524+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039888.6255388, "update": 18000} +{"device": "cuda", "elapsed_hours": 4.994725624630001, "elapsed_seconds": 17981.012248668005, "env_steps_per_second": 4115.4523992653285, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4049777533483785, "eta_seconds": 1457.9199120541625, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.003000194928785671, "chaser_grad_norm": 0.2113197147846222, "chaser_new_fields": 94.05000305175781, "chaser_partner_vision": 0.0, "chaser_return": 8.755000114440918, "chaser_subspace_norm": 85.910400390625, "collisions_per_episode": 1.649999976158142, "entropy": 0.0032143587595783174, "explorer_grad_norm": 0.3075653165578842, "explorer_new_fields": 94.07500457763672, "explorer_partner_vision": 0.0, "explorer_return": 90.8125, "explorer_subspace_norm": 129.43612670898438, "final_distance": 3.690819263458252, "max_param_abs": 2.547708511352539, "policy_loss": -0.001248595781159878, "value_loss": 0.1989212930202484}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:59:41.209364+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040381.2093768, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.129313266408345, "elapsed_seconds": 18465.527759070043, "env_steps_per_second": 4115.7773009043685, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.26996385612675505, "eta_seconds": 971.8698820563181, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": -9.281285058064526e-05, "chaser_grad_norm": 0.1693040239624679, "chaser_new_fields": 95.375, "chaser_partner_vision": 0.0, "chaser_return": 9.03499984741211, "chaser_subspace_norm": 87.02962493896484, "collisions_per_episode": 0.30000001192092896, "entropy": 0.0014391170989256352, "explorer_grad_norm": 0.26788607239723206, "explorer_new_fields": 95.45000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 129.20558166503906, "final_distance": 4.1876349449157715, "max_param_abs": 2.599325656890869, "policy_loss": -5.9722901816550156e-05, "value_loss": 0.04067488666623831}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:07:45.724874+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040865.7248878, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.2669596803119445, "elapsed_seconds": 18961.054849123, "env_steps_per_second": 4113.695183135221, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13505024821312678, "eta_seconds": 486.18089356725636, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 5.0394625173533214e-05, "chaser_grad_norm": 0.016865029465407133, "chaser_new_fields": 95.625, "chaser_partner_vision": 0.0, "chaser_return": 9.089999198913574, "chaser_subspace_norm": 83.32659912109375, "collisions_per_episode": 0.45000001788139343, "entropy": 0.003003425197675824, "explorer_grad_norm": 0.22870483994483948, "explorer_new_fields": 95.4749984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.0250015258789, "explorer_subspace_norm": 130.46420288085938, "final_distance": 4.799615859985352, "max_param_abs": 2.652148723602295, "policy_loss": -0.00012371969623181656, "value_loss": 0.054840755648911}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:16:01.251963+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041361.2519763, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.393116781728329, "elapsed_seconds": 19415.220414221985, "env_steps_per_second": 4120.478588097749, "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-transformer/non_social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:35.417527+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041815.4175391, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.3931393438688895, "elapsed_seconds": 19415.301637928, "env_steps_per_second": 4120.461350119801, "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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:35.498752+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041815.498765, "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..11f18a9aa66580c355a4de1f7565af86adb3d050 --- /dev/null +++ b/non_social/seed_0009/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.3931393438688895, + "elapsed_seconds": 19415.301637928, + "env_steps_per_second": 4120.461350119801, + "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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.002462393811179686, + "chaser_grad_norm": 0.06417040340602398, + "chaser_new_fields": 95.80000305175781, + "chaser_partner_vision": 0.0, + "chaser_return": 9.135000228881836, + "chaser_subspace_norm": 84.29200744628906, + "collisions_per_episode": 0.5, + "entropy": 0.002156060712877661, + "explorer_grad_norm": 0.33139311894774437, + "explorer_new_fields": 95.5999984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 93.17500305175781, + "explorer_subspace_norm": 135.6640625, + "final_distance": 4.808191776275635, + "max_param_abs": 2.7070395946502686, + "policy_loss": -0.0008346782419828003, + "value_loss": 0.06533607095479965 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:35.498752+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041815.498765, + "update": 20000 +} diff --git a/paper_random_eval.jsonl b/paper_random_eval.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..5252aad4732bfb224987b5ffefb20ded9f94b311 --- /dev/null +++ b/paper_random_eval.jsonl @@ -0,0 +1,40 @@ +{"attempt_id": "attempt_01", "batch_size": 128, "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.030178+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.1368327140808105, "chaser_new_fields": 93.50999450683594, "chaser_partner_vision": 0.0, "chaser_return": 8.491996765136719, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.059999942779541, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.11000061035156, "explorer_partner_vision": 0.0, "explorer_return": -0.949999988079071, "explorer_stuck_run_steps": 2.7200000286102295, "final_distance": 4.908422946929932, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0003983886290532812, "chaser_grad_norm": 0.02647677343338728, "chaser_new_fields": 94.9000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.955000877380371, "chaser_subspace_norm": 75.5232925415039, "collisions_per_episode": 0.550000011920929, "entropy": 0.0030606171931140125, "explorer_grad_norm": 0.17282858304679394, "explorer_new_fields": 95.42500305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.95000457763672, "explorer_subspace_norm": 155.51156616210938, "final_distance": 3.808248281478882, "max_param_abs": 2.2150280475616455, "policy_loss": -0.00044627454370527175, "value_loss": 0.06406432762742043}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.168256+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.048107147216797, "chaser_new_fields": 33.82999801635742, "chaser_partner_vision": 0.0, "chaser_return": -3.289998769760132, "chaser_stuck_run_steps": 2.609999895095825, "collisions_per_episode": 3.059999942779541, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 93.15999603271484, "explorer_partner_vision": 0.0, "explorer_return": 88.68999481201172, "explorer_stuck_run_steps": 0.0, "final_distance": 3.929388999938965, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.333558+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.137263298034668, "chaser_new_fields": 92.8699951171875, "chaser_partner_vision": 0.0, "chaser_return": 8.371996879577637, "chaser_stuck_run_steps": 0.029999999329447746, "collisions_per_episode": 3.3499999046325684, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.189998626708984, "explorer_partner_vision": 0.0, "explorer_return": -0.8899999856948853, "explorer_stuck_run_steps": 2.75, "final_distance": 4.113985061645508, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0031268233263772593, "chaser_grad_norm": 0.2260504849255085, "chaser_new_fields": 93.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.640000343322754, "chaser_subspace_norm": 81.0290298461914, "collisions_per_episode": 1.149999976158142, "entropy": 0.003422852954827249, "explorer_grad_norm": 0.24985111132264137, "explorer_new_fields": 94.82500457763672, "explorer_partner_vision": 0.0, "explorer_return": 91.82500457763672, "explorer_subspace_norm": 135.97732543945312, "final_distance": 2.2879505157470703, "max_param_abs": 3.1244027614593506, "policy_loss": -0.0035095140663869984, "value_loss": 0.2153652273118496}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.470364+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.043961048126221, "chaser_new_fields": 33.89999771118164, "chaser_partner_vision": 0.0, "chaser_return": -3.2999987602233887, "chaser_stuck_run_steps": 2.6399998664855957, "collisions_per_episode": 3.059999942779541, "degenerate_episodes": 89, "degenerate_fraction": 0.89, "episodes": 100, "explorer_new_fields": 93.68999481201172, "explorer_partner_vision": 0.0, "explorer_return": 89.37999725341797, "explorer_stuck_run_steps": 0.04999999701976776, "final_distance": 3.96620774269104, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.635312+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.096349239349365, "chaser_new_fields": 93.15999603271484, "chaser_partner_vision": 0.0, "chaser_return": 8.377996444702148, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.8999998569488525, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 33.07999801635742, "explorer_partner_vision": 0.0, "explorer_return": -1.2949999570846558, "explorer_stuck_run_steps": 2.740000009536743, "final_distance": 4.303606033325195, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 3.0426282704221208e-05, "chaser_grad_norm": 0.03582112770527601, "chaser_new_fields": 94.5250015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.820000648498535, "chaser_subspace_norm": 58.25376892089844, "collisions_per_episode": 0.9750000238418579, "entropy": 0.0015975794522091746, "explorer_grad_norm": 0.33543068170547485, "explorer_new_fields": 97.5250015258789, "explorer_partner_vision": 0.0, "explorer_return": 95.57500457763672, "explorer_subspace_norm": 127.81230163574219, "final_distance": 3.7922744750976562, "max_param_abs": 2.384674310684204, "policy_loss": -4.3445531147767724e-05, "value_loss": 0.07627437077462673}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.771588+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.124117374420166, "chaser_new_fields": 33.93000030517578, "chaser_partner_vision": 0.0, "chaser_return": -3.2579989433288574, "chaser_stuck_run_steps": 2.6499998569488525, "collisions_per_episode": 2.419999837875366, "degenerate_episodes": 90, "degenerate_fraction": 0.9, "episodes": 100, "explorer_new_fields": 95.88999938964844, "explorer_partner_vision": 0.0, "explorer_return": 93.05500030517578, "explorer_stuck_run_steps": 0.0, "final_distance": 3.837200880050659, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:11.936980+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.091452598571777, "chaser_new_fields": 92.63999938964844, "chaser_partner_vision": 0.0, "chaser_return": 8.311995506286621, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 3.3299999237060547, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.11000061035156, "explorer_partner_vision": 0.0, "explorer_return": -1.3249999284744263, "explorer_stuck_run_steps": 2.6499998569488525, "final_distance": 4.107217311859131, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0002613905893603663, "chaser_grad_norm": 0.02223971695639193, "chaser_new_fields": 93.7249984741211, "chaser_partner_vision": 0.0, "chaser_return": 8.6850004196167, "chaser_subspace_norm": 82.19728088378906, "collisions_per_episode": 1.0, "entropy": 0.0022482319036498666, "explorer_grad_norm": 0.44222569093108177, "explorer_new_fields": 95.7750015258789, "explorer_partner_vision": 0.0, "explorer_return": 93.0999984741211, "explorer_subspace_norm": 121.7215576171875, "final_distance": 2.864136219024658, "max_param_abs": 2.740300178527832, "policy_loss": -0.00019148159428716838, "value_loss": 0.08482817187905312}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.073269+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.068935871124268, "chaser_new_fields": 34.20000076293945, "chaser_partner_vision": 0.0, "chaser_return": -3.2079992294311523, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 2.7200000286102295, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 94.29999542236328, "explorer_partner_vision": 0.0, "explorer_return": 90.45999908447266, "explorer_stuck_run_steps": 0.0, "final_distance": 4.635434627532959, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.241084+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.109180927276611, "chaser_new_fields": 92.5199966430664, "chaser_partner_vision": 0.0, "chaser_return": 8.313996315002441, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 2.859999895095825, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 33.06999969482422, "explorer_partner_vision": 0.0, "explorer_return": -0.9799999594688416, "explorer_stuck_run_steps": 2.7200000286102295, "final_distance": 4.262806415557861, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0017689612679874234, "chaser_grad_norm": 0.025046652648597956, "chaser_new_fields": 94.0, "chaser_partner_vision": 0.0, "chaser_return": 8.77500057220459, "chaser_subspace_norm": 89.04515075683594, "collisions_per_episode": 0.4000000059604645, "entropy": 0.0025691293994896114, "explorer_grad_norm": 0.1489477176219225, "explorer_new_fields": 95.05000305175781, "explorer_partner_vision": 0.0, "explorer_return": 92.38750457763672, "explorer_subspace_norm": 140.36648559570312, "final_distance": 3.9162888526916504, "max_param_abs": 2.88704776763916, "policy_loss": -0.0007027814439490854, "value_loss": 0.05476926174014807}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.377530+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.085238456726074, "chaser_new_fields": 34.36000061035156, "chaser_partner_vision": 0.0, "chaser_return": -3.1899986267089844, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 2.9800000190734863, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 93.37999725341797, "explorer_partner_vision": 0.0, "explorer_return": 89.00499725341797, "explorer_stuck_run_steps": 0.0, "final_distance": 4.054072380065918, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.531673+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.154397487640381, "chaser_new_fields": 93.55999755859375, "chaser_partner_vision": 0.0, "chaser_return": 8.541996955871582, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 2.6599998474121094, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 32.939998626708984, "explorer_partner_vision": 0.0, "explorer_return": -1.1449999809265137, "explorer_stuck_run_steps": 2.759999990463257, "final_distance": 4.512866973876953, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 4.4411225102368446e-05, "chaser_grad_norm": 0.09301618672907352, "chaser_new_fields": 94.4000015258789, "chaser_partner_vision": 0.0, "chaser_return": 8.815000534057617, "chaser_subspace_norm": 69.54359436035156, "collisions_per_episode": 0.925000011920929, "entropy": 0.002813033468555659, "explorer_grad_norm": 0.6812433376908302, "explorer_new_fields": 97.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 96.0625, "explorer_subspace_norm": 156.21910095214844, "final_distance": 3.1009461879730225, "max_param_abs": 3.969390392303467, "policy_loss": -0.0003465476266564238, "value_loss": 0.07470510713756084}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.668574+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.100773334503174, "chaser_new_fields": 33.84000015258789, "chaser_partner_vision": 0.0, "chaser_return": -3.279998779296875, "chaser_stuck_run_steps": 2.5799999237060547, "collisions_per_episode": 2.5899999141693115, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 96.0, "explorer_partner_vision": 0.0, "explorer_return": 92.94999694824219, "explorer_stuck_run_steps": 0.009999999776482582, "final_distance": 4.033876895904541, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.832755+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.138570308685303, "chaser_new_fields": 92.82999420166016, "chaser_partner_vision": 0.0, "chaser_return": 8.367996215820312, "chaser_stuck_run_steps": 0.05999999865889549, "collisions_per_episode": 3.359999895095825, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.21999740600586, "explorer_partner_vision": 0.0, "explorer_return": -0.9350000023841858, "explorer_stuck_run_steps": 2.7100000381469727, "final_distance": 4.475497722625732, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.701445683692352e-05, "chaser_grad_norm": 0.033485319931060076, "chaser_new_fields": 94.57500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.830000877380371, "chaser_subspace_norm": 70.88212585449219, "collisions_per_episode": 0.9000000357627869, "entropy": 0.0015748643199913204, "explorer_grad_norm": 0.313498567789793, "explorer_new_fields": 97.55000305175781, "explorer_partner_vision": 0.0, "explorer_return": 95.80000305175781, "explorer_subspace_norm": 132.86404418945312, "final_distance": 2.430217742919922, "max_param_abs": 3.379441976547241, "policy_loss": -4.615307129274271e-05, "value_loss": 0.07251793146133423}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:12.968950+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.129418849945068, "chaser_new_fields": 33.880001068115234, "chaser_partner_vision": 0.0, "chaser_return": -3.269998788833618, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 2.319999933242798, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 96.38999938964844, "explorer_partner_vision": 0.0, "explorer_return": 93.50499725341797, "explorer_stuck_run_steps": 0.0, "final_distance": 3.9629032611846924, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.133560+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.0569047927856445, "chaser_new_fields": 90.43000030517578, "chaser_partner_vision": 0.0, "chaser_return": 7.805994510650635, "chaser_stuck_run_steps": 0.05999999865889549, "collisions_per_episode": 4.210000038146973, "degenerate_episodes": 84, "degenerate_fraction": 0.84, "episodes": 100, "explorer_new_fields": 33.05999755859375, "explorer_partner_vision": 0.0, "explorer_return": -1.3399999141693115, "explorer_stuck_run_steps": 2.690000057220459, "final_distance": 4.142829418182373, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 5.98452016989659e-05, "chaser_grad_norm": 0.023093817057088017, "chaser_new_fields": 91.82500457763672, "chaser_partner_vision": 0.0, "chaser_return": 8.360000610351562, "chaser_subspace_norm": 73.04069519042969, "collisions_per_episode": 0.3499999940395355, "entropy": 0.0009539475431665778, "explorer_grad_norm": 0.2525465562939644, "explorer_new_fields": 95.95000457763672, "explorer_partner_vision": 0.0, "explorer_return": 93.6624984741211, "explorer_subspace_norm": 149.63381958007812, "final_distance": 2.9168975353240967, "max_param_abs": 2.185072422027588, "policy_loss": -7.590103632537648e-05, "value_loss": 0.040869749151170254}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.270325+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.080101013183594, "chaser_new_fields": 34.290000915527344, "chaser_partner_vision": 0.0, "chaser_return": -3.1859986782073975, "chaser_stuck_run_steps": 2.6599998474121094, "collisions_per_episode": 2.919999837875366, "degenerate_episodes": 87, "degenerate_fraction": 0.87, "episodes": 100, "explorer_new_fields": 93.5199966430664, "explorer_partner_vision": 0.0, "explorer_return": 89.45499420166016, "explorer_stuck_run_steps": 0.009999999776482582, "final_distance": 4.23733377456665, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.435106+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.084146022796631, "chaser_new_fields": 93.05999755859375, "chaser_partner_vision": 0.0, "chaser_return": 8.329996109008789, "chaser_stuck_run_steps": 0.0, "collisions_per_episode": 3.9700000286102295, "degenerate_episodes": 82, "degenerate_fraction": 0.82, "episodes": 100, "explorer_new_fields": 32.97999954223633, "explorer_partner_vision": 0.0, "explorer_return": -1.59499990940094, "explorer_stuck_run_steps": 2.6399998664855957, "final_distance": 4.020629405975342, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 6.500302986564566e-05, "chaser_grad_norm": 0.026715639047324657, "chaser_new_fields": 94.5, "chaser_partner_vision": 0.0, "chaser_return": 8.855000495910645, "chaser_subspace_norm": 64.1446304321289, "collisions_per_episode": 0.7250000238418579, "entropy": 0.0010687967587728053, "explorer_grad_norm": 0.17880532704293728, "explorer_new_fields": 97.625, "explorer_partner_vision": 0.0, "explorer_return": 95.875, "explorer_subspace_norm": 122.6617660522461, "final_distance": 2.5540859699249268, "max_param_abs": 2.898890733718872, "policy_loss": -0.00012717402295669977, "value_loss": 0.05584537237882614}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.571327+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.054930686950684, "chaser_new_fields": 33.72999954223633, "chaser_partner_vision": 0.0, "chaser_return": -3.3139989376068115, "chaser_stuck_run_steps": 2.690000057220459, "collisions_per_episode": 2.950000047683716, "degenerate_episodes": 91, "degenerate_fraction": 0.91, "episodes": 100, "explorer_new_fields": 94.97000122070312, "explorer_partner_vision": 0.0, "explorer_return": 91.67499542236328, "explorer_stuck_run_steps": 0.0, "final_distance": 3.9026472568511963, "same_state_run_steps": 0.0}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.744621+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 5.109021186828613, "chaser_new_fields": 93.91999816894531, "chaser_partner_vision": 0.0, "chaser_return": 8.579997062683105, "chaser_stuck_run_steps": 0.009999999776482582, "collisions_per_episode": 3.25, "degenerate_episodes": 86, "degenerate_fraction": 0.86, "episodes": 100, "explorer_new_fields": 33.209999084472656, "explorer_partner_vision": 0.0, "explorer_return": -1.0399999618530273, "explorer_stuck_run_steps": 2.7100000381469727, "final_distance": 5.0355939865112305, "same_state_run_steps": 0.0}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002462393811179686, "chaser_grad_norm": 0.06417040340602398, "chaser_new_fields": 95.80000305175781, "chaser_partner_vision": 0.0, "chaser_return": 9.135000228881836, "chaser_subspace_norm": 84.29200744628906, "collisions_per_episode": 0.5, "entropy": 0.002156060712877661, "explorer_grad_norm": 0.33139311894774437, "explorer_new_fields": 95.5999984741211, "explorer_partner_vision": 0.0, "explorer_return": 93.17500305175781, "explorer_subspace_norm": 135.6640625, "final_distance": 4.808191776275635, "max_param_abs": 2.7070395946502686, "policy_loss": -0.0008346782419828003, "value_loss": 0.06533607095479965}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:13.881030+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 5.061388969421387, "chaser_new_fields": 34.25, "chaser_partner_vision": 0.0, "chaser_return": -3.2059991359710693, "chaser_stuck_run_steps": 2.6399998664855957, "collisions_per_episode": 2.950000047683716, "degenerate_episodes": 90, "degenerate_fraction": 0.9, "episodes": 100, "explorer_new_fields": 92.62999725341797, "explorer_partner_vision": 0.0, "explorer_return": 87.93999481201172, "explorer_stuck_run_steps": 0.08999999612569809, "final_distance": 4.295379638671875, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.053956+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5451525449752808, "chaser_new_fields": 26.189998626708984, "chaser_partner_vision": 0.9635999798774719, "chaser_return": 42.83700180053711, "chaser_stuck_run_steps": 0.07999999821186066, "collisions_per_episode": 43.98999786376953, "degenerate_episodes": 99, "degenerate_fraction": 0.99, "episodes": 100, "explorer_new_fields": 23.189998626708984, "explorer_partner_vision": 0.9635999798774719, "explorer_return": -44.709999084472656, "explorer_stuck_run_steps": 3.7799999713897705, "final_distance": 1.3304944038391113, "same_state_run_steps": 0.04999999701976776}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.198234+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.476309776306152, "chaser_new_fields": 34.02000045776367, "chaser_partner_vision": 0.17139999568462372, "chaser_return": -2.9389989376068115, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.25, "degenerate_episodes": 95, "degenerate_fraction": 0.95, "episodes": 100, "explorer_new_fields": 30.420000076293945, "explorer_partner_vision": 0.17139999568462372, "explorer_return": -4.630000114440918, "explorer_stuck_run_steps": 3.75, "final_distance": 6.77753210067749, "same_state_run_steps": 0.2800000011920929}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.371819+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.588372826576233, "chaser_new_fields": 27.989999771118164, "chaser_partner_vision": 0.9638999700546265, "chaser_return": 36.32500076293945, "chaser_stuck_run_steps": 0.3499999940395355, "collisions_per_episode": 37.72999954223633, "degenerate_episodes": 96, "degenerate_fraction": 0.96, "episodes": 100, "explorer_new_fields": 23.729999542236328, "explorer_partner_vision": 0.9638999700546265, "explorer_return": -40.91999816894531, "explorer_stuck_run_steps": 3.4099998474121094, "final_distance": 1.5478826761245728, "same_state_run_steps": 0.07999999821186066}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.516274+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.472661018371582, "chaser_new_fields": 33.97999954223633, "chaser_partner_vision": 0.17270000278949738, "chaser_return": -2.8129992485046387, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.3700000047683716, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 36.34000015258789, "explorer_partner_vision": 0.17270000278949738, "explorer_return": 4.159999847412109, "explorer_stuck_run_steps": 6.259999752044678, "final_distance": 6.965867042541504, "same_state_run_steps": 0.6599999666213989}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.695114+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5360196828842163, "chaser_new_fields": 26.75, "chaser_partner_vision": 0.9630999565124512, "chaser_return": 44.46000289916992, "chaser_stuck_run_steps": 0.3499999940395355, "collisions_per_episode": 45.36000061035156, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 23.31999969482422, "explorer_partner_vision": 0.9630999565124512, "explorer_return": -45.91999816894531, "explorer_stuck_run_steps": 3.6499998569488525, "final_distance": 1.4446792602539062, "same_state_run_steps": 0.07000000029802322}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:14.839830+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.161418914794922, "chaser_new_fields": 34.11000061035156, "chaser_partner_vision": 0.19539998471736908, "chaser_return": -2.780999183654785, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.3700000047683716, "degenerate_episodes": 93, "degenerate_fraction": 0.93, "episodes": 100, "explorer_new_fields": 43.939998626708984, "explorer_partner_vision": 0.19539998471736908, "explorer_return": 15.484999656677246, "explorer_stuck_run_steps": 3.3299999237060547, "final_distance": 6.8849663734436035, "same_state_run_steps": 0.26999998092651367}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.012176+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.4947947263717651, "chaser_new_fields": 26.31999969482422, "chaser_partner_vision": 0.9711999893188477, "chaser_return": 43.392005920410156, "chaser_stuck_run_steps": 0.12999999523162842, "collisions_per_episode": 44.439998626708984, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 22.15999984741211, "explorer_partner_vision": 0.9711999893188477, "explorer_return": -46.93000030517578, "explorer_stuck_run_steps": 3.7200000286102295, "final_distance": 1.3213845491409302, "same_state_run_steps": 0.009999999776482582}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.157169+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.446540355682373, "chaser_new_fields": 34.09000015258789, "chaser_partner_vision": 0.17020000517368317, "chaser_return": -2.8699989318847656, "chaser_stuck_run_steps": 2.6399998664855957, "collisions_per_episode": 0.29999998211860657, "degenerate_episodes": 98, "degenerate_fraction": 0.98, "episodes": 100, "explorer_new_fields": 46.5, "explorer_partner_vision": 0.17020000517368317, "explorer_return": 19.404998779296875, "explorer_stuck_run_steps": 6.899999618530273, "final_distance": 6.924410343170166, "same_state_run_steps": 0.5199999809265137}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.315288+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5326578617095947, "chaser_new_fields": 25.40999984741211, "chaser_partner_vision": 0.961899995803833, "chaser_return": 42.152000427246094, "chaser_stuck_run_steps": 0.14000000059604645, "collisions_per_episode": 43.459999084472656, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 22.09000015258789, "explorer_partner_vision": 0.961899995803833, "explorer_return": -45.70499801635742, "explorer_stuck_run_steps": 3.9800000190734863, "final_distance": 1.2944831848144531, "same_state_run_steps": 0.07000000029802322}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.459441+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.819681167602539, "chaser_new_fields": 34.09000015258789, "chaser_partner_vision": 0.13769999146461487, "chaser_return": -2.916998863220215, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.25, "degenerate_episodes": 95, "degenerate_fraction": 0.95, "episodes": 100, "explorer_new_fields": 36.32999801635742, "explorer_partner_vision": 0.13769999146461487, "explorer_return": 4.190000057220459, "explorer_stuck_run_steps": 7.349999904632568, "final_distance": 6.673317909240723, "same_state_run_steps": 0.699999988079071}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.617311+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.6489869356155396, "chaser_new_fields": 28.6299991607666, "chaser_partner_vision": 0.9557999968528748, "chaser_return": 36.21500015258789, "chaser_stuck_run_steps": 0.19999998807907104, "collisions_per_episode": 37.46999740600586, "degenerate_episodes": 96, "degenerate_fraction": 0.96, "episodes": 100, "explorer_new_fields": 24.939998626708984, "explorer_partner_vision": 0.9557999968528748, "explorer_return": -38.09000015258789, "explorer_stuck_run_steps": 3.2899999618530273, "final_distance": 1.4506957530975342, "same_state_run_steps": 0.05999999865889549}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.761441+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.512670040130615, "chaser_new_fields": 33.98999786376953, "chaser_partner_vision": 0.15479999780654907, "chaser_return": -2.978998899459839, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.20999999344348907, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 41.18000030517578, "explorer_partner_vision": 0.15479999780654907, "explorer_return": 11.529999732971191, "explorer_stuck_run_steps": 3.419999837875366, "final_distance": 6.647017478942871, "same_state_run_steps": 0.38999998569488525}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:15.918824+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.667794942855835, "chaser_new_fields": 27.760000228881836, "chaser_partner_vision": 0.9440000057220459, "chaser_return": 36.88300323486328, "chaser_stuck_run_steps": 0.8299999833106995, "collisions_per_episode": 38.290000915527344, "degenerate_episodes": 98, "degenerate_fraction": 0.98, "episodes": 100, "explorer_new_fields": 25.18000030517578, "explorer_partner_vision": 0.9440000057220459, "explorer_return": -38.994998931884766, "explorer_stuck_run_steps": 3.5999999046325684, "final_distance": 1.4672372341156006, "same_state_run_steps": 0.35999998450279236}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.062945+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.227531433105469, "chaser_new_fields": 34.02000045776367, "chaser_partner_vision": 0.20059996843338013, "chaser_return": -2.903998851776123, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.2800000011920929, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 41.709999084472656, "explorer_partner_vision": 0.20059996843338013, "explorer_return": 12.244999885559082, "explorer_stuck_run_steps": 5.0, "final_distance": 6.640198707580566, "same_state_run_steps": 0.38999998569488525}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.220422+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5710211992263794, "chaser_new_fields": 25.19999885559082, "chaser_partner_vision": 0.9613999724388123, "chaser_return": 40.952999114990234, "chaser_stuck_run_steps": 0.19999998807907104, "collisions_per_episode": 42.369998931884766, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 21.5, "explorer_partner_vision": 0.9613999724388123, "explorer_return": -46.85499954223633, "explorer_stuck_run_steps": 3.7200000286102295, "final_distance": 1.3192702531814575, "same_state_run_steps": 0.05999999865889549}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.364863+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.704348087310791, "chaser_new_fields": 33.939998626708984, "chaser_partner_vision": 0.15079998970031738, "chaser_return": -2.963999032974243, "chaser_stuck_run_steps": 2.629999876022339, "collisions_per_episode": 0.23999999463558197, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 30.35999870300293, "explorer_partner_vision": 0.15079998970031738, "explorer_return": -4.670000076293945, "explorer_stuck_run_steps": 15.00999927520752, "final_distance": 6.317955017089844, "same_state_run_steps": 1.3399999141693115}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.522956+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5945370197296143, "chaser_new_fields": 26.549999237060547, "chaser_partner_vision": 0.9594999551773071, "chaser_return": 34.46500015258789, "chaser_stuck_run_steps": 0.08999999612569809, "collisions_per_episode": 36.189998626708984, "degenerate_episodes": 99, "degenerate_fraction": 0.99, "episodes": 100, "explorer_new_fields": 22.969999313354492, "explorer_partner_vision": 0.9594999551773071, "explorer_return": -40.5099983215332, "explorer_stuck_run_steps": 3.7200000286102295, "final_distance": 1.4907904863357544, "same_state_run_steps": 0.03999999910593033}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.667220+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.724240779876709, "chaser_new_fields": 33.96999740600586, "chaser_partner_vision": 0.13729999959468842, "chaser_return": -2.9429991245269775, "chaser_stuck_run_steps": 2.5899999141693115, "collisions_per_episode": 0.25, "degenerate_episodes": 97, "degenerate_fraction": 0.97, "episodes": 100, "explorer_new_fields": 34.59000015258789, "explorer_partner_vision": 0.13729999959468842, "explorer_return": 1.5499999523162842, "explorer_stuck_run_steps": 14.15999984741211, "final_distance": 6.654120445251465, "same_state_run_steps": 1.0299999713897705}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.825633+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-transformer", "focus_agent": "chaser", "max_steps": 100, "metrics": {"average_distance": 1.5335965156555176, "chaser_new_fields": 27.619998931884766, "chaser_partner_vision": 0.964900016784668, "chaser_return": 42.8380012512207, "chaser_stuck_run_steps": 0.35999998450279236, "collisions_per_episode": 43.779998779296875, "degenerate_episodes": 100, "degenerate_fraction": 1.0, "episodes": 100, "explorer_new_fields": 24.18000030517578, "explorer_partner_vision": 0.964900016784668, "explorer_return": -44.54499816894531, "explorer_stuck_run_steps": 3.5199999809265137, "final_distance": 1.3675272464752197, "same_state_run_steps": 0.08999999612569809}, "opponent_mode": "random_explorer", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "checkpoint_config": {"architecture": "transformer", "attempt_id": "attempt_01", "batch_size": 40, "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", "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-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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": 266}, "checkpoint_metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "command": "/root/mouse-run-run-perf/.venv/bin/evaluate-paper-marl runs/mouse-run-run-2-transformer --output runs/mouse-run-run-2-transformer/paper_random_eval.jsonl --device cuda", "created_at": "2026-07-03T01:24:16.970033+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-transformer", "focus_agent": "explorer", "max_steps": 100, "metrics": {"average_distance": 6.735382080078125, "chaser_new_fields": 34.09000015258789, "chaser_partner_vision": 0.14399999380111694, "chaser_return": -2.9249989986419678, "chaser_stuck_run_steps": 2.619999885559082, "collisions_per_episode": 0.25, "degenerate_episodes": 94, "degenerate_fraction": 0.94, "episodes": 100, "explorer_new_fields": 34.63999938964844, "explorer_partner_vision": 0.14399999380111694, "explorer_return": 1.6999999284744263, "explorer_stuck_run_steps": 9.229999542236328, "final_distance": 6.582851409912109, "same_state_run_steps": 0.5699999928474426}, "opponent_mode": "random_chaser", "run_id": "mouse-run-run-2-transformer.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-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0d8f24a39c5d356b4621d62bc7265ae034017e22 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:79523bf93d01ed70b7e868b54bd70459c210b81fefc7e06836c12becd787418b +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..9ffb38cf7bc2279b043390936ff2fb6bf3986f92 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4a525f188e61798e466bfab07962c00ba3a32419728ab58c132854d5c15021a6 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..8267379f59fbfb127df5a134a3b8b49df325d2d5 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:1a3ffcae843364d6a41194d3fa074e3e38567f536a4c773bfce6d18933683747 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..4b9c36a610da673fcb300b97540cf319108e45c8 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fdff410f6f4486b5465b7d7e9b0eea876e81ec84939a13d65c18966e634c34f7 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..feeabcec4cd506719970adeef7d6ba9ca336d97b --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:8a30011c7be774107fd2f2ec56a1411497995644a45e40ae9042e69fe31b8a7d +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..a0864f48f53e4428ca531833839e6f64c0d4ccf6 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:331de7d69aa252a1e6c25f85652611cfe8c65e76dc2d65502eea41dba4370686 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..cc3e00d20298e02b31e49791b84e9d1050828593 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:49399764286403a10871d4a229bf4c66cc34b98d57dbf844493500003cff4e74 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..5c2f69a24830af4181df8e59a188bde51e6ccfdf --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:bdd9bfaf3009fd86f8b51becb610eac87ef0e29815d79f8a036fc7c78907441a +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b44d22bcbcefc4abb90373e838e58d93979ab96c --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:c46462eb90ada39824c0444c9e0be4903798ac5be414d9cf2d6edf098253bb45 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..7ee98fae2ea5137b467b940f4d904e7810bf5570 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:2f95b021dd1d9ea44cf0be2a67663e3e14a847b79713c15072c335dd1ee53f51 +size 46506715 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b21f2dc00093cd3f8834da6a8ee8a252f5f7de1b --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7b692d07d996c19666bcec9d3cbd1ca49895bbdaf265f987065c8c0a07a5bf67 +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..11983878cd75420a61a2aea30db8709518debf02 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c982d42a21462cc56fbabd3d2e3cd0d0367cf46aad3e7751355630c4c16a448 +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..86759a5ecc24b5978f63b7bca53142b6209ca412 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fbd829a179d3930a2e2023754d754bfcd5da946d6a364fee2c4c15764e81cf5a +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..c6c7da3153bc693ca13e434546ec3cb9539304d1 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7850ba0b8693746b2fc891d045f690281838a55ff69f0ca4e7b0792ffb600089 +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..0fe09caf6fcdb418466c8aa56056cc97fc20c0c5 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4d30fd2cf0d2050b4e36b40486d11328ffb2f6223eb365ed5d24adebf7a042d6 +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..be4a840d479f126f17a146c5d7c8d994447f68e1 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:61cab96377113bf461cc9b98496f068e46843299576259856285e4c5173396e4 +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..fe8868ac22ee014666ef992e0371f46941273235 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fd6b3b52c19b03cbe1bc347b56180787b658f72ee1d28efeddae9a403e23ca1d +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..b25256ac9252cdba25383bded6b7f7d2d53be3f9 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7ab97b49dbf08bc614e87382b28a0987ed96f4d5d3033b8b522164c2433de0fe +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..657320f9277f155a248741fac3bf823fdfed1170 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:efa669fe4caed034302ce55a56e212b2cc51232ab9140ebeaf18711ad69dabb0 +size 46506699 diff --git a/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..49d259d9ac6d6effbd2115481e5fc6bb97165cb3 --- /dev/null +++ b/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:148864f819405af7554f128fa3574e5d75c0bb6afc957d717e24cb5752490e52 +size 46506699 diff --git a/paper_rollouts/paper_rollout_records.jsonl b/paper_rollouts/paper_rollout_records.jsonl new file mode 100644 index 0000000000000000000000000000000000000000..1877c540cb7318b2ce4fb90b64d84cf177d8cac3 --- /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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:19.731502+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0000__attempt_01__checkpoints__latest__5edbca3078.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:20.616845+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0001__attempt_01__checkpoints__latest__00e1ed492a.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:21.500858+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0002__attempt_01__checkpoints__latest__97d42ca454.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:22.386345+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0003__attempt_01__checkpoints__latest__2e04f3c794.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:23.279182+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0004__attempt_01__checkpoints__latest__e4945dacc1.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:24.164102+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0005__attempt_01__checkpoints__latest__0c5efcab6f.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:25.051469+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0006__attempt_01__checkpoints__latest__333dda86d7.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:25.937613+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0007__attempt_01__checkpoints__latest__a49e9f5838.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:26.826889+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0008__attempt_01__checkpoints__latest__19594ed0fd.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:27.714387+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__non_social__seed_0009__attempt_01__checkpoints__latest__7662b25a03.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:28.637667+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0000__attempt_01__checkpoints__latest__1d93d9ea91.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:29.559684+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0001__attempt_01__checkpoints__latest__47857a85cd.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:30.484944+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0002__attempt_01__checkpoints__latest__5b27759942.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:31.410494+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0003__attempt_01__checkpoints__latest__70a7f3b121.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:32.334168+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0004__attempt_01__checkpoints__latest__5cbd564caf.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:33.266584+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0005__attempt_01__checkpoints__latest__9d297b2c34.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:34.189914+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0006__attempt_01__checkpoints__latest__d447912441.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:35.152778+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0007__attempt_01__checkpoints__latest__99724476e9.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:36.085812+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0008__attempt_01__checkpoints__latest__50b9fe7d0f.safetensors", "run_id": "mouse-run-run-2-transformer.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-transformer/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": 266}, "command": "/root/mouse-run-run-perf/.venv/bin/collect-paper-marl-rollouts runs/mouse-run-run-2-transformer --output-root runs/mouse-run-run-2-transformer/paper_rollouts --device cuda", "created_at": "2026-07-03T01:24:37.008624+00:00", "degenerate_threshold_fraction": 0.01, "deterministic": false, "device": "cuda", "episodes": 25, "error": null, "experiment_id": "mouse-run-run-2-transformer", "max_steps": 500, "output": "runs/mouse-run-run-2-transformer/paper_rollouts/mouse-run-run-2-transformer__social__seed_0009__attempt_01__checkpoints__latest__57a4097bd7.safetensors", "run_id": "mouse-run-run-2-transformer.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..1f5dfd1a42394f3fa5f8254af9ac637770b7a371 --- /dev/null +++ b/raw_records.jsonl @@ -0,0 +1,41 @@ +{"event": "runner_invocation", "experiment": "mouse-run-run-2-transformer", "payload": {"config": {"architecture": "transformer", "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": "6e419bd117b1aae2fbbef85bee470cef165112bab56d10a81e5dfbdc5bb6f9c5", "created_at": "2026-07-02T14:31:50.715150+00:00", "experiment": "mouse-run-run-2-transformer", "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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0000_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0000_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0001_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0001_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0002_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0002_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0003_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0003_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0004_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0004_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0005_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0005_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0006_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0006_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0007_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0007_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0008_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0008_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0009_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0009_attempt_01.log", "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.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_transformer.json", "created_at": "2026-07-02T14:31:50.715276+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": "2026-07-02T14:31:50.788465+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0000_attempt_01.log", "pid": 54459, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T14:31:50.789924+00:00", "started_monotonic": 652545.263907841, "task": "social", "unit_id": "social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.789930+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0000_attempt_01.log", "pid": 54460, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T14:31:50.790470+00:00", "started_monotonic": 652545.264456886, "task": "non_social", "unit_id": "non_social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.790480+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0001_attempt_01.log", "pid": 54461, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T14:31:50.791019+00:00", "started_monotonic": 652545.265006546, "task": "social", "unit_id": "social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.791028+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0001_attempt_01.log", "pid": 54462, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T14:31:50.791460+00:00", "started_monotonic": 652545.265442759, "task": "non_social", "unit_id": "non_social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.791464+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0002_attempt_01.log", "pid": 54463, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T14:31:50.791936+00:00", "started_monotonic": 652545.265918607, "task": "social", "unit_id": "social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.791940+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0002_attempt_01.log", "pid": 54464, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T14:31:50.792364+00:00", "started_monotonic": 652545.266347199, "task": "non_social", "unit_id": "non_social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.792368+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0003_attempt_01.log", "pid": 54465, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T14:31:50.792794+00:00", "started_monotonic": 652545.266777014, "task": "social", "unit_id": "social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.792798+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0003_attempt_01.log", "pid": 54466, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T14:31:50.793206+00:00", "started_monotonic": 652545.267188811, "task": "non_social", "unit_id": "non_social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.793210+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0004_attempt_01.log", "pid": 54467, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T14:31:50.793671+00:00", "started_monotonic": 652545.267654347, "task": "social", "unit_id": "social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.793676+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0004_attempt_01.log", "pid": 54468, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T14:31:50.794426+00:00", "started_monotonic": 652545.268423888, "task": "non_social", "unit_id": "non_social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T14:31:50.794456+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0000_attempt_01.log", "pid": 54460, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T14:31:50.790470+00:00", "started_monotonic": 652545.264456886, "task": "non_social", "unit_id": "non_social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:34.705093+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0003_attempt_01.log", "pid": 54466, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T14:31:50.793206+00:00", "started_monotonic": 652545.267188811, "task": "non_social", "unit_id": "non_social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:34.739282+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0005_attempt_01.log", "pid": 55092, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T19:57:44.742638+00:00", "started_monotonic": 672099.216629498, "task": "social", "unit_id": "social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:44.742659+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0005_attempt_01.log", "pid": 55093, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T19:57:44.743486+00:00", "started_monotonic": 672099.217474042, "task": "non_social", "unit_id": "non_social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:44.743498+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0002_attempt_01.log", "pid": 54464, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T14:31:50.792364+00:00", "started_monotonic": 652545.266347199, "task": "non_social", "unit_id": "non_social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:45.229752+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0006_attempt_01.log", "pid": 55160, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T19:57:55.232853+00:00", "started_monotonic": 672109.706841397, "task": "social", "unit_id": "social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:55.232866+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0003_attempt_01.log", "pid": 54465, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", "seed": 3, "started_at": "2026-07-02T14:31:50.792794+00:00", "started_monotonic": 652545.266777014, "task": "social", "unit_id": "social/seed_0003"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:56.185766+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0004_attempt_01.log", "pid": 54468, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T14:31:50.794426+00:00", "started_monotonic": 652545.268423888, "task": "non_social", "unit_id": "non_social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T19:57:57.287207+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0006_attempt_01.log", "pid": 55194, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T19:58:07.289631+00:00", "started_monotonic": 672121.763616238, "task": "non_social", "unit_id": "non_social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:07.289639+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0007_attempt_01.log", "pid": 55195, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T19:58:07.290258+00:00", "started_monotonic": 672121.764245459, "task": "social", "unit_id": "social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:07.290270+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0001_attempt_01.log", "pid": 54461, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T14:31:50.791019+00:00", "started_monotonic": 652545.265006546, "task": "social", "unit_id": "social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:07.307448+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0002_attempt_01.log", "pid": 54463, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", "seed": 2, "started_at": "2026-07-02T14:31:50.791936+00:00", "started_monotonic": 652545.265918607, "task": "social", "unit_id": "social/seed_0002"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:07.323530+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0007_attempt_01.log", "pid": 55262, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T19:58:17.326095+00:00", "started_monotonic": 672131.800084025, "task": "non_social", "unit_id": "non_social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:17.326108+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0008_attempt_01.log", "pid": 55263, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T19:58:17.326844+00:00", "started_monotonic": 672131.800831069, "task": "social", "unit_id": "social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:17.326854+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0004_attempt_01.log", "pid": 54467, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", "seed": 4, "started_at": "2026-07-02T14:31:50.793671+00:00", "started_monotonic": 652545.267654347, "task": "social", "unit_id": "social/seed_0004"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:37.412058+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0008_attempt_01.log", "pid": 55330, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T19:58:47.415107+00:00", "started_monotonic": 672161.889093166, "task": "non_social", "unit_id": "non_social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-02T19:58:47.415117+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0001_attempt_01.log", "pid": 54462, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", "seed": 1, "started_at": "2026-07-02T14:31:50.791460+00:00", "started_monotonic": 652545.265442759, "task": "non_social", "unit_id": "non_social/seed_0001"}, "schema_version": 1, "timestamp": "2026-07-02T19:59:27.473741+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0009_attempt_01.log", "pid": 55364, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T19:59:37.477716+00:00", "started_monotonic": 672211.951705928, "task": "social", "unit_id": "social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-02T19:59:37.477732+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0000_attempt_01.log", "pid": 54459, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", "seed": 0, "started_at": "2026-07-02T14:31:50.789924+00:00", "started_monotonic": 652545.263907841, "task": "social", "unit_id": "social/seed_0000"}, "schema_version": 1, "timestamp": "2026-07-02T19:59:49.335173+00:00"} +{"event": "attempt_started", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0009_attempt_01.log", "pid": 55398, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T19:59:59.339698+00:00", "started_monotonic": 672233.813687719, "task": "non_social", "unit_id": "non_social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-02T19:59:59.339713+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0006_attempt_01.log", "pid": 55194, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T19:58:07.289631+00:00", "started_monotonic": 672121.763616238, "task": "non_social", "unit_id": "non_social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-03T01:22:37.263784+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0006_attempt_01.log", "pid": 55160, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", "seed": 6, "started_at": "2026-07-02T19:57:55.232853+00:00", "started_monotonic": 672109.706841397, "task": "social", "unit_id": "social/seed_0006"}, "schema_version": 1, "timestamp": "2026-07-03T01:22:48.640779+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0005_attempt_01.log", "pid": 55093, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T19:57:44.743486+00:00", "started_monotonic": 672099.217474042, "task": "non_social", "unit_id": "non_social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:18.673750+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0008_attempt_01.log", "pid": 55263, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T19:58:17.326844+00:00", "started_monotonic": 672131.800831069, "task": "social", "unit_id": "social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:18.698125+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0005_attempt_01.log", "pid": 55092, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", "seed": 5, "started_at": "2026-07-02T19:57:44.742638+00:00", "started_monotonic": 672099.216629498, "task": "social", "unit_id": "social/seed_0005"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:28.725920+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0008_attempt_01.log", "pid": 55330, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", "seed": 8, "started_at": "2026-07-02T19:58:47.415107+00:00", "started_monotonic": 672161.889093166, "task": "non_social", "unit_id": "non_social/seed_0008"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:28.750751+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0009_attempt_01.log", "pid": 55398, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T19:59:59.339698+00:00", "started_monotonic": 672233.813687719, "task": "non_social", "unit_id": "non_social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:38.769580+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0007_attempt_01.log", "pid": 55195, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T19:58:07.290258+00:00", "started_monotonic": 672121.764245459, "task": "social", "unit_id": "social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:48.783609+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0007_attempt_01.log", "pid": 55262, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", "seed": 7, "started_at": "2026-07-02T19:58:17.326095+00:00", "started_monotonic": 672131.800084025, "task": "non_social", "unit_id": "non_social/seed_0007"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:48.795352+00:00"} +{"event": "attempt_completed", "experiment": "mouse-run-run-2-transformer", "payload": {"attempt": 1, "attempt_id": "attempt_01", "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "tensor_count": 266}, "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0009_attempt_01.log", "pid": 55364, "resume_from": null, "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", "seed": 9, "started_at": "2026-07-02T19:59:37.477716+00:00", "started_monotonic": 672211.951705928, "task": "social", "unit_id": "social/seed_0009"}, "schema_version": 1, "timestamp": "2026-07-03T01:23:58.814946+00:00"} diff --git a/run_status.json b/run_status.json new file mode 100644 index 0000000000000000000000000000000000000000..49069aa3d65bbb525e3dd8ff3c06e35eeeba685f --- /dev/null +++ b/run_status.json @@ -0,0 +1,989 @@ +{ + "completed_jobs": 20, + "concurrency": 10, + "elapsed_seconds": 39138.103445254965, + "experiment": "mouse-run-run-2-transformer", + "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": 5.463973339191095, + "elapsed_seconds": 19670.30402108794, + "env_steps_per_second": 4067.04440939166, + "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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0000_attempt_01.log", + "metrics": { + "approx_kl": 0.0030988860332676893, + "chaser_grad_norm": 0.36728256940841675, + "chaser_new_fields": 62.0, + "chaser_partner_vision": 0.9725000262260437, + "chaser_return": 13.717500686645508, + "chaser_subspace_norm": 272.79351806640625, + "collisions_per_episode": 10.925000190734863, + "entropy": 0.11317355558276176, + "explorer_grad_norm": 0.8740078210830688, + "explorer_new_fields": 69.95000457763672, + "explorer_partner_vision": 0.9725000262260437, + "explorer_return": 46.98749923706055, + "explorer_subspace_norm": 102.8924560546875, + "final_distance": 1.6796330213546753, + "max_param_abs": 2.4855704307556152, + "policy_loss": -0.0027665531197555993, + "value_loss": 1.877354085445404 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0000.attempt_01", + "schema_version": 2, + "seed": 0, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T19:59:42.252081+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0000", + "unix_time": 1783022382.2520971, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.427951933531375, + "elapsed_seconds": 19540.62696071295, + "env_steps_per_second": 4094.034452468825, + "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-transformer/non_social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0000_attempt_01.log", + "metrics": { + "approx_kl": 0.0003983886290532812, + "chaser_grad_norm": 0.02647677343338728, + "chaser_new_fields": 94.9000015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.955000877380371, + "chaser_subspace_norm": 75.5232925415039, + "collisions_per_episode": 0.550000011920929, + "entropy": 0.0030606171931140125, + "explorer_grad_norm": 0.17282858304679394, + "explorer_new_fields": 95.42500305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 92.95000457763672, + "explorer_subspace_norm": 155.51156616210938, + "final_distance": 3.808248281478882, + "max_param_abs": 2.2150280475616455, + "policy_loss": -0.00044627454370527175, + "value_loss": 0.06406432762742043 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0000.attempt_01", + "schema_version": 2, + "seed": 0, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T19:57:32.702201+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0000", + "unix_time": 1783022252.702217, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.435685596511111, + "elapsed_seconds": 19568.46814744, + "env_steps_per_second": 4088.2096338473903, + "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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0001_attempt_01.log", + "metrics": { + "approx_kl": 0.0032518112735848703, + "chaser_grad_norm": 0.4799686074256897, + "chaser_new_fields": 58.82500076293945, + "chaser_partner_vision": 0.9635000228881836, + "chaser_return": 13.932499885559082, + "chaser_subspace_norm": 267.9047546386719, + "collisions_per_episode": 11.824999809265137, + "entropy": 0.16130381077528, + "explorer_grad_norm": 1.6371502578258514, + "explorer_new_fields": 63.85000228881836, + "explorer_partner_vision": 0.9635000228881836, + "explorer_return": 35.51250076293945, + "explorer_subspace_norm": 96.23020935058594, + "final_distance": 1.6419010162353516, + "max_param_abs": 2.923971652984619, + "policy_loss": -0.0037321077841312444, + "value_loss": 3.0047545433044434 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0001.attempt_01", + "schema_version": 2, + "seed": 1, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T19:58:00.522236+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0001", + "unix_time": 1783022280.5222507, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.457300321458897, + "elapsed_seconds": 19646.28115725203, + "env_steps_per_second": 4072.0174652733, + "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-transformer/non_social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0001_attempt_01.log", + "metrics": { + "approx_kl": 0.0031268233263772593, + "chaser_grad_norm": 0.2260504849255085, + "chaser_new_fields": 93.4000015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.640000343322754, + "chaser_subspace_norm": 81.0290298461914, + "collisions_per_episode": 1.149999976158142, + "entropy": 0.003422852954827249, + "explorer_grad_norm": 0.24985111132264137, + "explorer_new_fields": 94.82500457763672, + "explorer_partner_vision": 0.0, + "explorer_return": 91.82500457763672, + "explorer_subspace_norm": 135.97732543945312, + "final_distance": 2.2879505157470703, + "max_param_abs": 3.1244027614593506, + "policy_loss": -0.0035095140663869984, + "value_loss": 0.2153652273118496 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0001.attempt_01", + "schema_version": 2, + "seed": 1, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T19:59:18.217854+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0001", + "unix_time": 1783022358.2178695, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.435672476867783, + "elapsed_seconds": 19568.42091672402, + "env_steps_per_second": 4088.2195012285606, + "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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0002_attempt_01.log", + "metrics": { + "approx_kl": 0.0025767051566676843, + "chaser_grad_norm": 0.482593297958374, + "chaser_new_fields": 56.32500076293945, + "chaser_partner_vision": 0.9642500281333923, + "chaser_return": 15.255000114440918, + "chaser_subspace_norm": 223.2849578857422, + "collisions_per_episode": 13.199999809265137, + "entropy": 0.16332043334841728, + "explorer_grad_norm": 0.6213180869817734, + "explorer_new_fields": 62.5, + "explorer_partner_vision": 0.9642500281333923, + "explorer_return": 33.775001525878906, + "explorer_subspace_norm": 86.63322448730469, + "final_distance": 1.4801658391952515, + "max_param_abs": 2.83324933052063, + "policy_loss": -0.004363851029381749, + "value_loss": 3.7607646584510803 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0002.attempt_01", + "schema_version": 2, + "seed": 2, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T19:58:00.355640+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0002", + "unix_time": 1783022280.355656, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.431030371984137, + "elapsed_seconds": 19551.709339142893, + "env_steps_per_second": 4091.713855414088, + "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-transformer/non_social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0002_attempt_01.log", + "metrics": { + "approx_kl": 3.0426282704221208e-05, + "chaser_grad_norm": 0.03582112770527601, + "chaser_new_fields": 94.5250015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.820000648498535, + "chaser_subspace_norm": 58.25376892089844, + "collisions_per_episode": 0.9750000238418579, + "entropy": 0.0015975794522091746, + "explorer_grad_norm": 0.33543068170547485, + "explorer_new_fields": 97.5250015258789, + "explorer_partner_vision": 0.0, + "explorer_return": 95.57500457763672, + "explorer_subspace_norm": 127.81230163574219, + "final_distance": 3.7922744750976562, + "max_param_abs": 2.384674310684204, + "policy_loss": -4.3445531147767724e-05, + "value_loss": 0.07627437077462673 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0002.attempt_01", + "schema_version": 2, + "seed": 2, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T19:57:43.888084+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0002", + "unix_time": 1783022263.8880963, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.432549700080546, + "elapsed_seconds": 19557.178920289967, + "env_steps_per_second": 4090.569520586759, + "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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0003_attempt_01.log", + "metrics": { + "approx_kl": 0.002856620680569577, + "chaser_grad_norm": 0.7219458222389221, + "chaser_new_fields": 60.275001525878906, + "chaser_partner_vision": 0.9662500619888306, + "chaser_return": 13.242500305175781, + "chaser_subspace_norm": 225.597900390625, + "collisions_per_episode": 10.875, + "entropy": 0.1406782865524292, + "explorer_grad_norm": 0.60902339220047, + "explorer_new_fields": 70.80000305175781, + "explorer_partner_vision": 0.9662500619888306, + "explorer_return": 47.42499923706055, + "explorer_subspace_norm": 101.78258514404297, + "final_distance": 1.8629791736602783, + "max_param_abs": 2.9420251846313477, + "policy_loss": -0.005349479220825515, + "value_loss": 2.428272843360901 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0003.attempt_01", + "schema_version": 2, + "seed": 3, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T19:57:49.062689+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0003", + "unix_time": 1783022269.0627043, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.427600972299988, + "elapsed_seconds": 19539.363500279956, + "env_steps_per_second": 4094.2991822048743, + "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-transformer/non_social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0003_attempt_01.log", + "metrics": { + "approx_kl": 0.0002613905893603663, + "chaser_grad_norm": 0.02223971695639193, + "chaser_new_fields": 93.7249984741211, + "chaser_partner_vision": 0.0, + "chaser_return": 8.6850004196167, + "chaser_subspace_norm": 82.19728088378906, + "collisions_per_episode": 1.0, + "entropy": 0.0022482319036498666, + "explorer_grad_norm": 0.44222569093108177, + "explorer_new_fields": 95.7750015258789, + "explorer_partner_vision": 0.0, + "explorer_return": 93.0999984741211, + "explorer_subspace_norm": 121.7215576171875, + "final_distance": 2.864136219024658, + "max_param_abs": 2.740300178527832, + "policy_loss": -0.00019148159428716838, + "value_loss": 0.08482817187905312 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0003.attempt_01", + "schema_version": 2, + "seed": 3, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T19:57:31.531925+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0003", + "unix_time": 1783022251.531941, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.444820641110548, + "elapsed_seconds": 19601.35430799797, + "env_steps_per_second": 4081.350642560319, + "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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0004_attempt_01.log", + "metrics": { + "approx_kl": 0.004475183506503555, + "chaser_grad_norm": 0.6177457571029663, + "chaser_new_fields": 59.85000228881836, + "chaser_partner_vision": 0.9817500710487366, + "chaser_return": 9.820000648498535, + "chaser_subspace_norm": 235.4537811279297, + "collisions_per_episode": 7.550000190734863, + "entropy": 0.16257280483841896, + "explorer_grad_norm": 1.1996710449457169, + "explorer_new_fields": 68.70000457763672, + "explorer_partner_vision": 0.9817500710487366, + "explorer_return": 47.10000228881836, + "explorer_subspace_norm": 95.60823822021484, + "final_distance": 1.5751407146453857, + "max_param_abs": 2.528623104095459, + "policy_loss": -0.005616524885397212, + "value_loss": 2.955764353275299 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0004.attempt_01", + "schema_version": 2, + "seed": 4, + "state": "completed", + "task": "social", + "timestamp": "2026-07-02T19:58:33.222594+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0004", + "unix_time": 1783022313.2226105, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.433294409401101, + "elapsed_seconds": 19559.85987384396, + "env_steps_per_second": 4090.0088505735375, + "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-transformer/non_social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0004_attempt_01.log", + "metrics": { + "approx_kl": 0.0017689612679874234, + "chaser_grad_norm": 0.025046652648597956, + "chaser_new_fields": 94.0, + "chaser_partner_vision": 0.0, + "chaser_return": 8.77500057220459, + "chaser_subspace_norm": 89.04515075683594, + "collisions_per_episode": 0.4000000059604645, + "entropy": 0.0025691293994896114, + "explorer_grad_norm": 0.1489477176219225, + "explorer_new_fields": 95.05000305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 92.38750457763672, + "explorer_subspace_norm": 140.36648559570312, + "final_distance": 3.9162888526916504, + "max_param_abs": 2.88704776763916, + "policy_loss": -0.0007027814439490854, + "value_loss": 0.05476926174014807 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0004.attempt_01", + "schema_version": 2, + "seed": 4, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-02T19:57:51.852322+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0004", + "unix_time": 1783022271.8523347, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.4258453104955455, + "elapsed_seconds": 19533.043117783964, + "env_steps_per_second": 4095.623990465857, + "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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0005_attempt_01.log", + "metrics": { + "approx_kl": 0.003451461161873226, + "chaser_grad_norm": 0.4533055052161217, + "chaser_new_fields": 55.75, + "chaser_partner_vision": 0.9477500319480896, + "chaser_return": 9.160000801086426, + "chaser_subspace_norm": 194.28756713867188, + "collisions_per_episode": 7.650000095367432, + "entropy": 0.20668550580739975, + "explorer_grad_norm": 0.7168894857168198, + "explorer_new_fields": 63.625, + "explorer_partner_vision": 0.9477500319480896, + "explorer_return": 39.0625, + "explorer_subspace_norm": 89.7675552368164, + "final_distance": 1.5833224058151245, + "max_param_abs": 2.8546435832977295, + "policy_loss": -0.007046535808569843, + "value_loss": 3.2317103147506714 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0005.attempt_01", + "schema_version": 2, + "seed": 5, + "state": "completed", + "task": "social", + "timestamp": "2026-07-03T01:23:18.846686+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0005", + "unix_time": 1783041798.8466992, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.423259738547204, + "elapsed_seconds": 19523.735058769933, + "env_steps_per_second": 4097.5766040250855, + "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-transformer/non_social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0005_attempt_01.log", + "metrics": { + "approx_kl": 4.4411225102368446e-05, + "chaser_grad_norm": 0.09301618672907352, + "chaser_new_fields": 94.4000015258789, + "chaser_partner_vision": 0.0, + "chaser_return": 8.815000534057617, + "chaser_subspace_norm": 69.54359436035156, + "collisions_per_episode": 0.925000011920929, + "entropy": 0.002813033468555659, + "explorer_grad_norm": 0.6812433376908302, + "explorer_new_fields": 97.95000457763672, + "explorer_partner_vision": 0.0, + "explorer_return": 96.0625, + "explorer_subspace_norm": 156.21910095214844, + "final_distance": 3.1009461879730225, + "max_param_abs": 3.969390392303467, + "policy_loss": -0.0003465476266564238, + "value_loss": 0.07470510713756084 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0005.attempt_01", + "schema_version": 2, + "seed": 5, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-03T01:23:09.573312+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0005", + "unix_time": 1783041789.573325, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.411338810608334, + "elapsed_seconds": 19480.819718190003, + "env_steps_per_second": 4106.603374872407, + "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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0006_attempt_01.log", + "metrics": { + "approx_kl": 0.0015035698563394817, + "chaser_grad_norm": 0.5757317766547203, + "chaser_new_fields": 63.22500228881836, + "chaser_partner_vision": 0.9730000495910645, + "chaser_return": 12.269999504089355, + "chaser_subspace_norm": 190.25613403320312, + "collisions_per_episode": 9.300000190734863, + "entropy": 0.18287497758865356, + "explorer_grad_norm": 0.6242441534996033, + "explorer_new_fields": 69.2249984741211, + "explorer_partner_vision": 0.9730000495910645, + "explorer_return": 46.0, + "explorer_subspace_norm": 97.26615905761719, + "final_distance": 1.7034721374511719, + "max_param_abs": 2.8255860805511475, + "policy_loss": -0.005794699959466243, + "value_loss": 2.655023217201233 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0006.attempt_01", + "schema_version": 2, + "seed": 6, + "state": "completed", + "task": "social", + "timestamp": "2026-07-03T01:22:37.513633+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0006", + "unix_time": 1783041757.5136476, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.406713302494997, + "elapsed_seconds": 19464.16788898199, + "env_steps_per_second": 4110.116623340745, + "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-transformer/non_social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0006_attempt_01.log", + "metrics": { + "approx_kl": 5.701445683692352e-05, + "chaser_grad_norm": 0.033485319931060076, + "chaser_new_fields": 94.57500457763672, + "chaser_partner_vision": 0.0, + "chaser_return": 8.830000877380371, + "chaser_subspace_norm": 70.88212585449219, + "collisions_per_episode": 0.9000000357627869, + "entropy": 0.0015748643199913204, + "explorer_grad_norm": 0.313498567789793, + "explorer_new_fields": 97.55000305175781, + "explorer_partner_vision": 0.0, + "explorer_return": 95.80000305175781, + "explorer_subspace_norm": 132.86404418945312, + "final_distance": 2.430217742919922, + "max_param_abs": 3.379441976547241, + "policy_loss": -4.615307129274271e-05, + "value_loss": 0.07251793146133423 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0006.attempt_01", + "schema_version": 2, + "seed": 6, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-03T01:22:32.342773+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0006", + "unix_time": 1783041752.3427885, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.425122272546384, + "elapsed_seconds": 19530.44018116698, + "env_steps_per_second": 4096.169838360492, + "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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0007_attempt_01.log", + "metrics": { + "approx_kl": 0.004598107310812227, + "chaser_grad_norm": 0.545984223484993, + "chaser_new_fields": 57.54999923706055, + "chaser_partner_vision": 0.9745000600814819, + "chaser_return": 13.77500057220459, + "chaser_subspace_norm": 288.3200988769531, + "collisions_per_episode": 11.90000057220459, + "entropy": 0.1392931044101715, + "explorer_grad_norm": 0.9233212769031525, + "explorer_new_fields": 65.2249984741211, + "explorer_partner_vision": 0.9745000600814819, + "explorer_return": 38.17499923706055, + "explorer_subspace_norm": 111.23908996582031, + "final_distance": 1.6038007736206055, + "max_param_abs": 2.665212392807007, + "policy_loss": -0.005061323422069108, + "value_loss": 3.0026434659957886 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0007.attempt_01", + "schema_version": 2, + "seed": 7, + "state": "completed", + "task": "social", + "timestamp": "2026-07-03T01:23:38.578376+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0007", + "unix_time": 1783041818.5783887, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.424772208754698, + "elapsed_seconds": 19529.17995151691, + "env_steps_per_second": 4096.4341666474365, + "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-transformer/non_social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0007_attempt_01.log", + "metrics": { + "approx_kl": 5.98452016989659e-05, + "chaser_grad_norm": 0.023093817057088017, + "chaser_new_fields": 91.82500457763672, + "chaser_partner_vision": 0.0, + "chaser_return": 8.360000610351562, + "chaser_subspace_norm": 73.04069519042969, + "collisions_per_episode": 0.3499999940395355, + "entropy": 0.0009539475431665778, + "explorer_grad_norm": 0.2525465562939644, + "explorer_new_fields": 95.95000457763672, + "explorer_partner_vision": 0.0, + "explorer_return": 93.6624984741211, + "explorer_subspace_norm": 149.63381958007812, + "final_distance": 2.9168975353240967, + "max_param_abs": 2.185072422027588, + "policy_loss": -7.590103632537648e-05, + "value_loss": 0.040869749151170254 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0007.attempt_01", + "schema_version": 2, + "seed": 7, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-03T01:23:47.568405+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0007", + "unix_time": 1783041827.5684185, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.416619449098329, + "elapsed_seconds": 19499.830016753986, + "env_steps_per_second": 4102.599865294472, + "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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0008_attempt_01.log", + "metrics": { + "approx_kl": 0.003203999427618953, + "chaser_grad_norm": 1.5083578824996948, + "chaser_new_fields": 57.400001525878906, + "chaser_partner_vision": 0.9695000648498535, + "chaser_return": 16.552501678466797, + "chaser_subspace_norm": 264.9256591796875, + "collisions_per_episode": 14.475000381469727, + "entropy": 0.1569875180721283, + "explorer_grad_norm": 0.9279506653547287, + "explorer_new_fields": 64.4000015258789, + "explorer_partner_vision": 0.9695000648498535, + "explorer_return": 34.45000076293945, + "explorer_subspace_norm": 103.04727935791016, + "final_distance": 1.7443459033966064, + "max_param_abs": 2.8595657348632812, + "policy_loss": -0.004643450012451922, + "value_loss": 2.9393985867500305 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0008.attempt_01", + "schema_version": 2, + "seed": 8, + "state": "completed", + "task": "social", + "timestamp": "2026-07-03T01:23:18.090398+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0008", + "unix_time": 1783041798.0904107, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.408702007005275, + "elapsed_seconds": 19471.327225218993, + "env_steps_per_second": 4108.605390616882, + "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-transformer/non_social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0008_attempt_01.log", + "metrics": { + "approx_kl": 6.500302986564566e-05, + "chaser_grad_norm": 0.026715639047324657, + "chaser_new_fields": 94.5, + "chaser_partner_vision": 0.0, + "chaser_return": 8.855000495910645, + "chaser_subspace_norm": 64.1446304321289, + "collisions_per_episode": 0.7250000238418579, + "entropy": 0.0010687967587728053, + "explorer_grad_norm": 0.17880532704293728, + "explorer_new_fields": 97.625, + "explorer_partner_vision": 0.0, + "explorer_return": 95.875, + "explorer_subspace_norm": 122.6617660522461, + "final_distance": 2.5540859699249268, + "max_param_abs": 2.898890733718872, + "policy_loss": -0.00012717402295669977, + "value_loss": 0.05584537237882614 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0008.attempt_01", + "schema_version": 2, + "seed": 8, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-03T01:23:19.557206+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0008", + "unix_time": 1783041799.557218, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.404370285466947, + "elapsed_seconds": 19455.73302768101, + "env_steps_per_second": 4111.898528119115, + "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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/social_seed_0009_attempt_01.log", + "metrics": { + "approx_kl": 0.0012853747145520344, + "chaser_grad_norm": 1.286025196313858, + "chaser_new_fields": 57.650001525878906, + "chaser_partner_vision": 0.986750066280365, + "chaser_return": 11.644999504089355, + "chaser_subspace_norm": 265.3517761230469, + "collisions_per_episode": 9.75, + "entropy": 0.11431185901165009, + "explorer_grad_norm": 1.4314807802438736, + "explorer_new_fields": 67.17500305175781, + "explorer_partner_vision": 0.986750066280365, + "explorer_return": 41.72500228881836, + "explorer_subspace_norm": 115.21052551269531, + "final_distance": 1.8594802618026733, + "max_param_abs": 2.6002023220062256, + "policy_loss": -0.002069776682645852, + "value_loss": 2.937978982925415 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-transformer.social.seed_0009.attempt_01", + "schema_version": 2, + "seed": 9, + "state": "completed", + "task": "social", + "timestamp": "2026-07-03T01:23:54.118472+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "social/seed_0009", + "unix_time": 1783041834.1184835, + "update": 20000 + }, + { + "attempt": 1, + "attempt_id": "attempt_01", + "device": "cuda", + "elapsed_hours": 5.3931393438688895, + "elapsed_seconds": 19415.301637928, + "env_steps_per_second": 4120.461350119801, + "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-transformer/non_social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "log": "runs/mouse-run-run-2-transformer/logs/non_social_seed_0009_attempt_01.log", + "metrics": { + "approx_kl": 0.002462393811179686, + "chaser_grad_norm": 0.06417040340602398, + "chaser_new_fields": 95.80000305175781, + "chaser_partner_vision": 0.0, + "chaser_return": 9.135000228881836, + "chaser_subspace_norm": 84.29200744628906, + "collisions_per_episode": 0.5, + "entropy": 0.002156060712877661, + "explorer_grad_norm": 0.33139311894774437, + "explorer_new_fields": 95.5999984741211, + "explorer_partner_vision": 0.0, + "explorer_return": 93.17500305175781, + "explorer_subspace_norm": 135.6640625, + "final_distance": 4.808191776275635, + "max_param_abs": 2.7070395946502686, + "policy_loss": -0.0008346782419828003, + "value_loss": 0.06533607095479965 + }, + "progress_fraction": 1.0, + "run_dir": "runs/mouse-run-run-2-transformer/non_social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-transformer.non_social.seed_0009.attempt_01", + "schema_version": 2, + "seed": 9, + "state": "completed", + "task": "non_social", + "timestamp": "2026-07-03T01:23:35.498752+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unit_id": "non_social/seed_0009", + "unix_time": 1783041815.498765, + "update": 20000 + } + ], + "known_attempts": 20, + "observed_updates": 400000, + "pending_jobs": 0, + "progress_fraction": 1.0, + "running": [], + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:24:08.817697+00:00", + "total_jobs": 20, + "total_updates": 400000, + "unix_time": 1783041848.8177106, + "valid_pairs_by_task": { + "non_social": 10, + "social": 10 + } +} 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..f8101ed881616a2be7a8f6aab3aa2424ccfa83d3 --- /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:d43d74b10a625e6385b2bb95ea1788823a3302c478fc62063d73fa7275a2e64b +size 29760376 diff --git a/social/seed_0000/attempt_01/config.json b/social/seed_0000/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..daaa7d7284e15634e66d161c1f9b92627c889e1f --- /dev/null +++ b/social/seed_0000/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0000/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..38eece7a7656b497ff10bc6fa3c274a904ee93b2 --- /dev/null +++ b/social/seed_0000/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007575525772214556, "elapsed_seconds": 2.7271892779972404, "env_steps_per_second": 1466.7115452057917, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 15.150293991851893, "eta_seconds": 54541.05837066681, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004490051377276316, "chaser_grad_norm": 3.463568292558193, "chaser_new_fields": 35.07500076293945, "chaser_partner_vision": 0.3215000033378601, "chaser_return": -1.037500262260437, "chaser_subspace_norm": 0.03570802882313728, "collisions_per_episode": 1.8250000476837158, "entropy": 2.7544467449188232, "explorer_grad_norm": 6.090027809143066, "explorer_new_fields": 34.17499923706055, "explorer_partner_vision": 0.3215000033378601, "explorer_return": -0.13750000298023224, "explorer_subspace_norm": 0.04141807556152344, "final_distance": 5.855244159698486, "max_param_abs": 1.0012015104293823, "policy_loss": -0.014284036476155393, "value_loss": 7.896866083145142}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.675251+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.6752663, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13716985821471706, "elapsed_seconds": 493.8114895729814, "env_steps_per_second": 4050.1285252181565, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.349624470373965, "eta_seconds": 19258.648093346274, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0036013250703987865, "chaser_grad_norm": 0.8981616795063019, "chaser_new_fields": 59.42499923706055, "chaser_partner_vision": 0.4425000250339508, "chaser_return": 11.542499542236328, "chaser_subspace_norm": 10.626852989196777, "collisions_per_episode": 8.975000381469727, "entropy": 0.5510473549365997, "explorer_grad_norm": 2.0764722526073456, "explorer_new_fields": 65.2750015258789, "explorer_partner_vision": 0.4425000250339508, "explorer_return": 40.79999923706055, "explorer_subspace_norm": 7.421437740325928, "final_distance": 4.196762561798096, "max_param_abs": 1.1904914379119873, "policy_loss": -0.005090915353967773, "value_loss": 3.15396910905838}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:05.759552+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003205.7595658, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2748396687233213, "elapsed_seconds": 989.4228074039565, "env_steps_per_second": 4042.7610623765418, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.221953705743104, "eta_seconds": 18799.033340675174, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.003191894516179794, "chaser_grad_norm": 1.0868233889341354, "chaser_new_fields": 54.375, "chaser_partner_vision": 0.5412500500679016, "chaser_return": 12.120000839233398, "chaser_subspace_norm": 17.07150650024414, "collisions_per_episode": 10.5, "entropy": 0.4979819059371948, "explorer_grad_norm": 1.8905494213104248, "explorer_new_fields": 64.67500305175781, "explorer_partner_vision": 0.5412500500679016, "explorer_return": 38.837501525878906, "explorer_subspace_norm": 10.502326011657715, "final_distance": 3.6782569885253906, "max_param_abs": 1.2906389236450195, "policy_loss": -0.0056970714777762055, "value_loss": 3.591597616672516}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:21.370867+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003701.3708808, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.41387242347360975, "elapsed_seconds": 1489.9407245049952, "env_steps_per_second": 4027.0058407816105, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.104426556174521, "eta_seconds": 18375.935602228274, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0034061177545468, "chaser_grad_norm": 1.0257553160190582, "chaser_new_fields": 50.900001525878906, "chaser_partner_vision": 0.6150000095367432, "chaser_return": 12.72249984741211, "chaser_subspace_norm": 22.853744506835938, "collisions_per_episode": 11.824999809265137, "entropy": 0.38630010932683945, "explorer_grad_norm": 2.1558799147605896, "explorer_new_fields": 66.32500457763672, "explorer_partner_vision": 0.6150000095367432, "explorer_return": 39.41250228881836, "explorer_subspace_norm": 16.467924118041992, "final_distance": 3.3334596157073975, "max_param_abs": 1.3370938301086426, "policy_loss": -0.0061642665220489334, "value_loss": 3.4527416825294495}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:41.888783+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004201.8887973, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5515446790536064, "elapsed_seconds": 1985.560844592983, "env_steps_per_second": 4029.088316172909, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.963902111482457, "eta_seconds": 17870.047601336846, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.005174705976811467, "chaser_grad_norm": 1.261529564857483, "chaser_new_fields": 46.125, "chaser_partner_vision": 0.655750036239624, "chaser_return": 10.487500190734863, "chaser_subspace_norm": 27.64363670349121, "collisions_per_episode": 10.574999809265137, "entropy": 0.42658068984746933, "explorer_grad_norm": 1.298361375927925, "explorer_new_fields": 63.32500076293945, "explorer_partner_vision": 0.655750036239624, "explorer_return": 36.13750076293945, "explorer_subspace_norm": 16.68861198425293, "final_distance": 4.304214000701904, "max_param_abs": 1.3727437257766724, "policy_loss": -0.007758290320637684, "value_loss": 3.552015721797943}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:57.508905+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004697.5089188, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6870875205949738, "elapsed_seconds": 2473.5150741419056, "env_steps_per_second": 4042.82961706596, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.809612644164816, "eta_seconds": 17314.60551899334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0034622493465867876, "chaser_grad_norm": 1.432431936264038, "chaser_new_fields": 50.5, "chaser_partner_vision": 0.7105000615119934, "chaser_return": 13.347500801086426, "chaser_subspace_norm": 35.23208999633789, "collisions_per_episode": 12.324999809265137, "entropy": 0.3776041567325592, "explorer_grad_norm": 2.3116103410720825, "explorer_new_fields": 63.95000076293945, "explorer_partner_vision": 0.7105000615119934, "explorer_return": 35.48749923706055, "explorer_subspace_norm": 20.273853302001953, "final_distance": 3.0203194618225098, "max_param_abs": 1.4187835454940796, "policy_loss": -0.0036343382950860814, "value_loss": 3.8227829933166504}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:13:05.463134+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005185.4631495, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8260074350316426, "elapsed_seconds": 2973.6267661139136, "env_steps_per_second": 4035.476185762953, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.680708798512642, "eta_seconds": 16850.55167464551, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004696476328513555, "chaser_grad_norm": 1.0237432569265366, "chaser_new_fields": 51.10000228881836, "chaser_partner_vision": 0.780500054359436, "chaser_return": 15.030000686645508, "chaser_subspace_norm": 34.125980377197266, "collisions_per_episode": 13.90000057220459, "entropy": 0.4151345044374466, "explorer_grad_norm": 1.5701046586036682, "explorer_new_fields": 61.525001525878906, "explorer_partner_vision": 0.780500054359436, "explorer_return": 31.512500762939453, "explorer_subspace_norm": 21.50933074951172, "final_distance": 2.7726199626922607, "max_param_abs": 1.4298073053359985, "policy_loss": -0.007065867681063764, "value_loss": 3.7177592515945435}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:21:25.574827+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005685.574841, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9608349145344174, "elapsed_seconds": 3459.0056923239026, "env_steps_per_second": 4047.40588634135, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.529650311376539, "eta_seconds": 16306.74112095554, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.005445661835437932, "chaser_grad_norm": 0.9820783287286758, "chaser_new_fields": 44.92499923706055, "chaser_partner_vision": 0.8782500624656677, "chaser_return": 22.92750358581543, "chaser_subspace_norm": 35.97073745727539, "collisions_per_episode": 22.42500114440918, "entropy": 0.44802556931972504, "explorer_grad_norm": 1.4601117968559265, "explorer_new_fields": 58.17499923706055, "explorer_partner_vision": 0.8782500624656677, "explorer_return": 19.899999618530273, "explorer_subspace_norm": 22.76081085205078, "final_distance": 2.0662829875946045, "max_param_abs": 1.4708380699157715, "policy_loss": -0.007369668463255152, "value_loss": 4.727087378501892}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:30.953751+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006170.9537637, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0959619003666579, "elapsed_seconds": 3945.462841319968, "env_steps_per_second": 4055.291012358678, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.383847601466631, "eta_seconds": 15781.851365279872, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.005033102358851238, "chaser_grad_norm": 1.2986486703157425, "chaser_new_fields": 48.375, "chaser_partner_vision": 0.9287500381469727, "chaser_return": 26.125, "chaser_subspace_norm": 51.8447151184082, "collisions_per_episode": 24.75, "entropy": 0.3206745311617851, "explorer_grad_norm": 1.2831627428531647, "explorer_new_fields": 54.875, "explorer_partner_vision": 0.9287500381469727, "explorer_return": 13.487500190734863, "explorer_subspace_norm": 27.760761260986328, "final_distance": 2.0082690715789795, "max_param_abs": 1.5114986896514893, "policy_loss": -0.005898315721182712, "value_loss": 4.488527059555054}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:37.410902+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006657.4109185, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2339902323163632, "elapsed_seconds": 4442.3648363389075, "env_steps_per_second": 4051.895930014241, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.250410800200807, "eta_seconds": 15301.478880722905, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.006904012051691666, "chaser_grad_norm": 0.8644527345895767, "chaser_new_fields": 48.92499923706055, "chaser_partner_vision": 0.9597500562667847, "chaser_return": 19.122501373291016, "chaser_subspace_norm": 58.31239700317383, "collisions_per_episode": 18.17500114440918, "entropy": 0.3201195150613785, "explorer_grad_norm": 1.5671762973070145, "explorer_new_fields": 58.25, "explorer_partner_vision": 0.9597500562667847, "explorer_return": 22.212499618530273, "explorer_subspace_norm": 31.77205467224121, "final_distance": 1.849401831626892, "max_param_abs": 1.550491452217102, "policy_loss": -0.008906666411093056, "value_loss": 4.2013373374938965}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:54.312896+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007154.3129094, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3698483166238618, "elapsed_seconds": 4931.453939845902, "env_steps_per_second": 4055.5990675287458, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.109544949871585, "eta_seconds": 14794.361819537706, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.004643202841004168, "chaser_grad_norm": 0.5359167829155922, "chaser_new_fields": 50.20000076293945, "chaser_partner_vision": 0.9652500748634338, "chaser_return": 18.547500610351562, "chaser_subspace_norm": 61.38166809082031, "collisions_per_episode": 17.524999618530273, "entropy": 0.35546068847179413, "explorer_grad_norm": 1.0957041084766388, "explorer_new_fields": 56.57500076293945, "explorer_partner_vision": 0.9652500748634338, "explorer_return": 20.512500762939453, "explorer_subspace_norm": 27.883031845092773, "final_distance": 1.761924386024475, "max_param_abs": 1.5687243938446045, "policy_loss": -0.006318603320039529, "value_loss": 4.023233771324158}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:54:03.402000+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007643.4020143, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.5060094891316516, "elapsed_seconds": 5421.634160873946, "env_steps_per_second": 4057.81713542503, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9703886531652635, "eta_seconds": 14293.399151394948, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00631863320398196, "chaser_grad_norm": 0.774323433637619, "chaser_new_fields": 51.45000076293945, "chaser_partner_vision": 0.9515000581741333, "chaser_return": 18.627500534057617, "chaser_subspace_norm": 72.34402465820312, "collisions_per_episode": 17.42500114440918, "entropy": 0.2946373000741005, "explorer_grad_norm": 1.3476596474647522, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.9515000581741333, "explorer_return": 25.325000762939453, "explorer_subspace_norm": 33.00871658325195, "final_distance": 1.762311339378357, "max_param_abs": 1.580735206604004, "policy_loss": -0.007867062201725616, "value_loss": 3.966305673122406}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:02:13.582221+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008133.5822332, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6424160361741411, "elapsed_seconds": 5912.697730226908, "env_steps_per_second": 4059.0608712004914, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.832304084406329, "eta_seconds": 13796.294703862784, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0030812717019017555, "chaser_grad_norm": 0.4328109174966812, "chaser_new_fields": 49.650001525878906, "chaser_partner_vision": 0.955750048160553, "chaser_return": 19.475000381469727, "chaser_subspace_norm": 77.45746612548828, "collisions_per_episode": 18.5, "entropy": 0.2373313568532467, "explorer_grad_norm": 1.1211307495832443, "explorer_new_fields": 57.22500228881836, "explorer_partner_vision": 0.955750048160553, "explorer_return": 20.700000762939453, "explorer_subspace_norm": 39.88627624511719, "final_distance": 1.5298653841018677, "max_param_abs": 1.5839853286743164, "policy_loss": -0.0046468574795994755, "value_loss": 3.9930858612060547}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:10:24.645789+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008624.6458046, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7771682769535821, "elapsed_seconds": 6397.8057970328955, "env_steps_per_second": 4063.8932822965644, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6910418059805163, "eta_seconds": 13287.75050152986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.005444459672311552, "chaser_grad_norm": 0.5208474323153496, "chaser_new_fields": 54.07500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 16.315000534057617, "chaser_subspace_norm": 87.37273406982422, "collisions_per_episode": 14.75, "entropy": 0.22774096950888634, "explorer_grad_norm": 1.346103310585022, "explorer_new_fields": 61.17499923706055, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 28.950000762939453, "explorer_subspace_norm": 42.21167755126953, "final_distance": 1.598131537437439, "max_param_abs": 1.6129244565963745, "policy_loss": -0.007245788858284641, "value_loss": 3.378819227218628}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:18:29.753856+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009109.7538671, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.914126340347769, "elapsed_seconds": 6890.854825251969, "env_steps_per_second": 4063.3565370427264, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.554806060645857, "eta_seconds": 12797.301818325084, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.002202447284694209, "chaser_grad_norm": 1.6773909330368042, "chaser_new_fields": 52.75, "chaser_partner_vision": 0.9707500338554382, "chaser_return": 15.5, "chaser_subspace_norm": 99.81421661376953, "collisions_per_episode": 14.199999809265137, "entropy": 0.2558497115969658, "explorer_grad_norm": 2.6570059657096863, "explorer_new_fields": 61.45000076293945, "explorer_partner_vision": 0.9707500338554382, "explorer_return": 29.75, "explorer_subspace_norm": 43.18271255493164, "final_distance": 1.4896209239959717, "max_param_abs": 1.659415602684021, "policy_loss": -0.004267239525916011, "value_loss": 3.513906240463257}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:26:42.802885+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009602.8028994, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0508324920691665, "elapsed_seconds": 7382.996971449, "env_steps_per_second": 4063.3905331417395, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.418054153448611, "eta_seconds": 12304.994952415, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0018056226462377367, "chaser_grad_norm": 0.5359542965888977, "chaser_new_fields": 54.5, "chaser_partner_vision": 0.9690000414848328, "chaser_return": 14.27500057220459, "chaser_subspace_norm": 100.037353515625, "collisions_per_episode": 12.800000190734863, "entropy": 0.22411713376641273, "explorer_grad_norm": 1.203916296362877, "explorer_new_fields": 61.5, "explorer_partner_vision": 0.9690000414848328, "explorer_return": 31.837499618530273, "explorer_subspace_norm": 42.86784744262695, "final_distance": 1.5847705602645874, "max_param_abs": 1.691174030303955, "policy_loss": -0.004598108496992381, "value_loss": 3.2528817653656006}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:34:54.945031+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010094.945043, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1848166648727623, "elapsed_seconds": 7865.339993541944, "env_steps_per_second": 4068.482738988332, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.277224997309143, "eta_seconds": 11798.009990312916, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0023802689350418405, "chaser_grad_norm": 1.673940747976303, "chaser_new_fields": 50.17499923706055, "chaser_partner_vision": 0.955500066280365, "chaser_return": 17.5625, "chaser_subspace_norm": 115.93865966796875, "collisions_per_episode": 16.725000381469727, "entropy": 0.19799356535077095, "explorer_grad_norm": 1.1467385739088058, "explorer_new_fields": 54.97500228881836, "explorer_partner_vision": 0.955500066280365, "explorer_return": 19.67500114440918, "explorer_subspace_norm": 53.85546875, "final_distance": 1.3892384767532349, "max_param_abs": 1.7328470945358276, "policy_loss": -0.005077638575812671, "value_loss": 3.7645264267921448}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:57.288054+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010577.2880695, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.319891947543333, "elapsed_seconds": 8351.611011155997, "env_steps_per_second": 4071.070833469512, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1386773407939206, "eta_seconds": 11299.238426858114, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.005506305699597647, "chaser_grad_norm": 0.7098329812288284, "chaser_new_fields": 48.47500228881836, "chaser_partner_vision": 0.9467500448226929, "chaser_return": 17.567501068115234, "chaser_subspace_norm": 113.6498031616211, "collisions_per_episode": 16.67500114440918, "entropy": 0.23782162368297577, "explorer_grad_norm": 1.059070497751236, "explorer_new_fields": 55.22500228881836, "explorer_partner_vision": 0.9467500448226929, "explorer_return": 20.48750114440918, "explorer_subspace_norm": 55.57926559448242, "final_distance": 1.702405333518982, "max_param_abs": 1.7961573600769043, "policy_loss": -0.003277872466497911, "value_loss": 4.480886816978455}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:51:03.559072+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011063.559086, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.45617080931555, "elapsed_seconds": 8842.21491353598, "env_steps_per_second": 4071.3780825310987, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.0019865447190055, "eta_seconds": 10807.151560988419, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.002114475752350131, "chaser_grad_norm": 1.1742712259292603, "chaser_new_fields": 58.75, "chaser_partner_vision": 0.952250063419342, "chaser_return": 12.780000686645508, "chaser_subspace_norm": 111.46244049072266, "collisions_per_episode": 10.65000057220459, "entropy": 0.19867366179823875, "explorer_grad_norm": 1.0465290397405624, "explorer_new_fields": 66.375, "explorer_partner_vision": 0.952250063419342, "explorer_return": 40.26250076293945, "explorer_subspace_norm": 55.04779052734375, "final_distance": 1.6291977167129517, "max_param_abs": 1.8955144882202148, "policy_loss": -0.008215883376077926, "value_loss": 2.855012834072113}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:59:14.162974+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011554.162989, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.590493809927753, "elapsed_seconds": 9325.77771573991, "env_steps_per_second": 4074.7271871882767, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8631773688675164, "eta_seconds": 10307.43852792306, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.002701915020137946, "chaser_grad_norm": 0.42073817551136017, "chaser_new_fields": 57.20000076293945, "chaser_partner_vision": 0.9687500596046448, "chaser_return": 13.3774995803833, "chaser_subspace_norm": 121.37437438964844, "collisions_per_episode": 11.574999809265137, "entropy": 0.1664222963154316, "explorer_grad_norm": 1.2047697752714157, "explorer_new_fields": 65.42500305175781, "explorer_partner_vision": 0.9687500596046448, "explorer_return": 37.400001525878906, "explorer_subspace_norm": 60.66834259033203, "final_distance": 1.7249469757080078, "max_param_abs": 1.962226390838623, "policy_loss": -0.00454456434258077, "value_loss": 2.670404851436615}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:07:17.725774+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012037.7257862, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.727965972910816, "elapsed_seconds": 9820.677502478939, "env_steps_per_second": 4073.0387480805866, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.727965972910816, "eta_seconds": 9820.677502478939, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.003332214512715481, "chaser_grad_norm": 0.8886457085609436, "chaser_new_fields": 57.5, "chaser_partner_vision": 0.9552500247955322, "chaser_return": 12.522500038146973, "chaser_subspace_norm": 119.2752914428711, "collisions_per_episode": 10.574999809265137, "entropy": 0.19167208671569824, "explorer_grad_norm": 0.9187363982200623, "explorer_new_fields": 66.05000305175781, "explorer_partner_vision": 0.9552500247955322, "explorer_return": 39.88750076293945, "explorer_subspace_norm": 55.291019439697266, "final_distance": 1.7750362157821655, "max_param_abs": 1.9971801042556763, "policy_loss": -0.004251788744703333, "value_loss": 3.048218011856079}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:15:32.625562+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012532.625576, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.862935841756375, "elapsed_seconds": 10306.56903032295, "env_steps_per_second": 4075.0709451837783, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.590275285398625, "eta_seconds": 9324.99102743505, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.004143424316321642, "chaser_grad_norm": 0.7809827998280525, "chaser_new_fields": 50.95000076293945, "chaser_partner_vision": 0.9567500352859497, "chaser_return": 14.550000190734863, "chaser_subspace_norm": 132.2965850830078, "collisions_per_episode": 13.75, "entropy": 0.19856536388397217, "explorer_grad_norm": 0.8780129700899124, "explorer_new_fields": 58.22500228881836, "explorer_partner_vision": 0.9567500352859497, "explorer_return": 25.700000762939453, "explorer_subspace_norm": 64.10365295410156, "final_distance": 1.558486819267273, "max_param_abs": 2.0479629039764404, "policy_loss": -0.005979411132614132, "value_loss": 3.543888509273529}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:38.517089+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013018.5171, "update": 10500} +{"device": "cuda", "elapsed_hours": 3.000635410301103, "elapsed_seconds": 10802.28747708397, "env_steps_per_second": 4073.2113539231236, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4550653357009025, "eta_seconds": 8838.235208523249, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0021584276179087425, "chaser_grad_norm": 0.33995987102389336, "chaser_new_fields": 58.97500228881836, "chaser_partner_vision": 0.9565000534057617, "chaser_return": 11.710000991821289, "chaser_subspace_norm": 134.79165649414062, "collisions_per_episode": 9.550000190734863, "entropy": 0.17464479431509972, "explorer_grad_norm": 1.721743494272232, "explorer_new_fields": 67.92500305175781, "explorer_partner_vision": 0.9565000534057617, "explorer_return": 43.17499923706055, "explorer_subspace_norm": 61.4150276184082, "final_distance": 1.8465824127197266, "max_param_abs": 2.1484475135803223, "policy_loss": -0.0038654327345994943, "value_loss": 2.6395506262779236}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:54.235536+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013514.235552, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.136719334374971, "elapsed_seconds": 11292.189603749895, "env_steps_per_second": 4073.612081816655, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3184447254075873, "eta_seconds": 8346.401011467315, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.002882604546968537, "chaser_grad_norm": 0.4507957771420479, "chaser_new_fields": 56.54999923706055, "chaser_partner_vision": 0.9445000290870667, "chaser_return": 10.595000267028809, "chaser_subspace_norm": 129.37059020996094, "collisions_per_episode": 9.0, "entropy": 0.18501467630267143, "explorer_grad_norm": 1.5521484315395355, "explorer_new_fields": 63.400001525878906, "explorer_partner_vision": 0.9445000290870667, "explorer_return": 37.26250076293945, "explorer_subspace_norm": 71.36125183105469, "final_distance": 1.6467514038085938, "max_param_abs": 2.197619915008545, "policy_loss": -0.006070774896997477, "value_loss": 3.1286264061927795}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:40:04.137663+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014004.1376736, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.276428247350527, "elapsed_seconds": 11795.141690461896, "env_steps_per_second": 4069.472097890528, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1842854982336846, "eta_seconds": 7863.427793641264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.002158199512281911, "chaser_grad_norm": 0.7273614853620529, "chaser_new_fields": 49.82500076293945, "chaser_partner_vision": 0.9452500343322754, "chaser_return": 10.887499809265137, "chaser_subspace_norm": 128.75460815429688, "collisions_per_episode": 10.425000190734863, "entropy": 0.23467447236180305, "explorer_grad_norm": 1.0755289047956467, "explorer_new_fields": 55.82500076293945, "explorer_partner_vision": 0.9452500343322754, "explorer_return": 26.125, "explorer_subspace_norm": 60.5119514465332, "final_distance": 1.5083225965499878, "max_param_abs": 2.276460647583008, "policy_loss": -0.0060194242224085315, "value_loss": 3.5216992497444153}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:48:27.089749+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014507.0897636, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.414147247399443, "elapsed_seconds": 12290.930090637994, "env_steps_per_second": 4068.040386795871, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.048488348439666, "eta_seconds": 7374.558054382796, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0034110930049351396, "chaser_grad_norm": 0.5549731329083443, "chaser_new_fields": 49.375, "chaser_partner_vision": 0.95250004529953, "chaser_return": 12.307500839233398, "chaser_subspace_norm": 139.8497772216797, "collisions_per_episode": 11.77500057220459, "entropy": 0.21759935840964317, "explorer_grad_norm": 1.1088499277830124, "explorer_new_fields": 53.79999923706055, "explorer_partner_vision": 0.95250004529953, "explorer_return": 22.149999618530273, "explorer_subspace_norm": 66.36642456054688, "final_distance": 1.616832971572876, "max_param_abs": 2.2856011390686035, "policy_loss": -0.00477015676907655, "value_loss": 3.5476319789886475}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:56:42.878150+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015002.878161, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.551692959689438, "elapsed_seconds": 12786.094654881977, "env_steps_per_second": 4066.918117186423, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9124500552173898, "eta_seconds": 6884.820198782603, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.009835878360183958, "chaser_grad_norm": 4.752179384231567, "chaser_new_fields": 43.67499923706055, "chaser_partner_vision": 0.9602500200271606, "chaser_return": 6.367500305175781, "chaser_subspace_norm": 147.89727783203125, "collisions_per_episode": 7.275000095367432, "entropy": 0.1439991444349289, "explorer_grad_norm": 2.0492110550403595, "explorer_new_fields": 49.07500076293945, "explorer_partner_vision": 0.9602500200271606, "explorer_return": 17.5, "explorer_subspace_norm": 81.09732055664062, "final_distance": 1.4586246013641357, "max_param_abs": 2.2877883911132812, "policy_loss": 0.0009975887383575355, "value_loss": 4.364446640014648}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:04:58.042716+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015498.0427318, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6890198021358165, "elapsed_seconds": 13280.471287688939, "env_steps_per_second": 4066.1207595891765, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7761947195468748, "eta_seconds": 6394.300990368749, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0020651114816156446, "chaser_grad_norm": 0.7908197566866875, "chaser_new_fields": 56.875, "chaser_partner_vision": 0.9437500238418579, "chaser_return": 10.720000267028809, "chaser_subspace_norm": 166.93734741210938, "collisions_per_episode": 9.050000190734863, "entropy": 0.16169371083378792, "explorer_grad_norm": 1.0042861700057983, "explorer_new_fields": 65.95000457763672, "explorer_partner_vision": 0.9437500238418579, "explorer_return": 40.23749923706055, "explorer_subspace_norm": 68.46318817138672, "final_distance": 1.7424182891845703, "max_param_abs": 2.302194595336914, "policy_loss": -0.00574010954608184, "value_loss": 2.546234905719757}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:13:12.419346+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015992.4193573, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.82399308358248, "elapsed_seconds": 13766.375100896927, "env_steps_per_second": 4067.8827643125464, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6388541786782056, "eta_seconds": 5899.87504324154, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.002596058282483771, "chaser_grad_norm": 0.666592463850975, "chaser_new_fields": 58.85000228881836, "chaser_partner_vision": 0.9540000557899475, "chaser_return": 11.455000877380371, "chaser_subspace_norm": 159.36317443847656, "collisions_per_episode": 9.25, "entropy": 0.16870694234967232, "explorer_grad_norm": 0.9895422458648682, "explorer_new_fields": 65.80000305175781, "explorer_partner_vision": 0.9540000557899475, "explorer_return": 40.88750076293945, "explorer_subspace_norm": 75.07255554199219, "final_distance": 1.438172698020935, "max_param_abs": 2.313830852508545, "policy_loss": -0.005573441524063583, "value_loss": 3.1165120005607605}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:21:18.323163+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016478.3231783, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9613274113919275, "elapsed_seconds": 14260.77868101094, "env_steps_per_second": 4067.099090264292, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.5025724663900415, "eta_seconds": 5409.2608790041495, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.004737566865543386, "chaser_grad_norm": 0.5755189135670662, "chaser_new_fields": 49.47500228881836, "chaser_partner_vision": 0.9502500295639038, "chaser_return": 12.780001640319824, "chaser_subspace_norm": 168.8978271484375, "collisions_per_episode": 12.15000057220459, "entropy": 0.19009919837117195, "explorer_grad_norm": 0.7658346742391586, "explorer_new_fields": 55.400001525878906, "explorer_partner_vision": 0.9502500295639038, "explorer_return": 24.287500381469727, "explorer_subspace_norm": 70.58683776855469, "final_distance": 1.4853315353393555, "max_param_abs": 2.3233845233917236, "policy_loss": -0.004221188301244183, "value_loss": 3.845008671283722}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:29:32.726743+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016972.7267587, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.096948498976094, "elapsed_seconds": 14749.014596313937, "env_steps_per_second": 4068.0683857344043, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3656494996586979, "eta_seconds": 4916.338198771312, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.009356312464966265, "chaser_grad_norm": 0.5662543773651123, "chaser_new_fields": 48.04999923706055, "chaser_partner_vision": 0.9750000238418579, "chaser_return": 19.16499900817871, "chaser_subspace_norm": 138.07778930664062, "collisions_per_episode": 18.350000381469727, "entropy": 0.22392147406935692, "explorer_grad_norm": 0.879424437880516, "explorer_new_fields": 52.42499923706055, "explorer_partner_vision": 0.9750000238418579, "explorer_return": 16.3125, "explorer_subspace_norm": 81.6526870727539, "final_distance": 1.6477346420288086, "max_param_abs": 2.3713507652282715, "policy_loss": -0.004321216898034663, "value_loss": 4.5502811670303345}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:37:40.962655+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017460.9626698, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.233761051343042, "elapsed_seconds": 15241.539784834953, "env_steps_per_second": 4067.830473512187, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2291564342608834, "eta_seconds": 4424.96316333918, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0020912183440486842, "chaser_grad_norm": 0.6161137223243713, "chaser_new_fields": 52.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 22.065000534057617, "chaser_subspace_norm": 151.29319763183594, "collisions_per_episode": 20.399999618530273, "entropy": 0.22740261629223824, "explorer_grad_norm": 0.6890130043029785, "explorer_new_fields": 61.57500076293945, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 26.462499618530273, "explorer_subspace_norm": 66.68730926513672, "final_distance": 1.5314778089523315, "max_param_abs": 2.449155569076538, "policy_loss": -0.003799335133749082, "value_loss": 3.802286922931671}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:45:53.487846+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017953.4878595, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.370033505512757, "elapsed_seconds": 15732.120619845926, "env_steps_per_second": 4068.11017703898, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0925083763781893, "eta_seconds": 3933.0301549614815, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.002855258153977802, "chaser_grad_norm": 0.576439693570137, "chaser_new_fields": 56.72500228881836, "chaser_partner_vision": 0.9345000386238098, "chaser_return": 13.267500877380371, "chaser_subspace_norm": 172.0299530029297, "collisions_per_episode": 11.475000381469727, "entropy": 0.22796984389424324, "explorer_grad_norm": 1.000185251235962, "explorer_new_fields": 65.5, "explorer_partner_vision": 0.9345000386238098, "explorer_return": 39.17499923706055, "explorer_subspace_norm": 74.97310638427734, "final_distance": 1.519074559211731, "max_param_abs": 2.4549520015716553, "policy_loss": -0.005259015386911869, "value_loss": 3.3725106716156006}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:54:04.068681+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018444.068696, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.506644116802749, "elapsed_seconds": 16223.918820489896, "env_steps_per_second": 4068.0676925383596, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9559548126551286, "eta_seconds": 3441.437325558463, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0019146054615699626, "chaser_grad_norm": 0.6662779077887535, "chaser_new_fields": 60.32500076293945, "chaser_partner_vision": 0.9720000624656677, "chaser_return": 14.4975004196167, "chaser_subspace_norm": 195.71368408203125, "collisions_per_episode": 11.925000190734863, "entropy": 0.15702836960554123, "explorer_grad_norm": 0.8330029249191284, "explorer_new_fields": 66.875, "explorer_partner_vision": 0.9720000624656677, "explorer_return": 41.125, "explorer_subspace_norm": 82.21076202392578, "final_distance": 1.67135488986969, "max_param_abs": 2.44755220413208, "policy_loss": -0.0028385315984280624, "value_loss": 2.8607149720191956}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:02:15.866880+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018935.8668926, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.644570234859421, "elapsed_seconds": 16720.452845493914, "env_steps_per_second": 4066.8754984303964, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8196300414457801, "eta_seconds": 2950.6681492048083, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0020042440625758218, "chaser_grad_norm": 0.5851510837674141, "chaser_new_fields": 57.0, "chaser_partner_vision": 0.971500039100647, "chaser_return": 12.789999961853027, "chaser_subspace_norm": 198.68077087402344, "collisions_per_episode": 11.0, "entropy": 0.17697186022996902, "explorer_grad_norm": 0.6333399564027786, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.971500039100647, "explorer_return": 34.5, "explorer_subspace_norm": 74.46171569824219, "final_distance": 1.5103315114974976, "max_param_abs": 2.463942050933838, "policy_loss": -0.005342885947754894, "value_loss": 3.179632604122162}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:10:32.400905+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019432.400919, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.780190462594425, "elapsed_seconds": 17208.68566533993, "env_steps_per_second": 4067.713325776368, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6828843517992036, "eta_seconds": 2458.383666477133, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0020652039903978903, "chaser_grad_norm": 0.7073933482170105, "chaser_new_fields": 53.67499923706055, "chaser_partner_vision": 0.9777500629425049, "chaser_return": 16.08500099182129, "chaser_subspace_norm": 206.8367919921875, "collisions_per_episode": 14.699999809265137, "entropy": 0.1861644722521305, "explorer_grad_norm": 0.8900148123502731, "explorer_new_fields": 60.17499923706055, "explorer_partner_vision": 0.9777500629425049, "explorer_return": 28.86250114440918, "explorer_subspace_norm": 90.21548461914062, "final_distance": 1.4877080917358398, "max_param_abs": 2.45900559425354, "policy_loss": -0.004373061076279328, "value_loss": 3.33663272857666}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:18:40.633726+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019920.63374, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.917698427228606, "elapsed_seconds": 17703.714338022983, "env_steps_per_second": 4066.9431637496937, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5464109363587341, "eta_seconds": 1967.0793708914425, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0026511270188933622, "chaser_grad_norm": 0.4771323874592781, "chaser_new_fields": 56.650001525878906, "chaser_partner_vision": 0.9755000472068787, "chaser_return": 14.072501182556152, "chaser_subspace_norm": 209.79254150390625, "collisions_per_episode": 12.27500057220459, "entropy": 0.176283061504364, "explorer_grad_norm": 0.8287802934646606, "explorer_new_fields": 65.20000457763672, "explorer_partner_vision": 0.9755000472068787, "explorer_return": 37.837501525878906, "explorer_subspace_norm": 89.63745880126953, "final_distance": 1.5419012308120728, "max_param_abs": 2.4867420196533203, "policy_loss": -0.0043985866969507015, "value_loss": 3.2459129691123962}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:26:55.662397+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020415.662411, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.057838107955823, "elapsed_seconds": 18208.217188640963, "env_steps_per_second": 4064.0991500345376, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.41009498172614783, "eta_seconds": 1476.3419342141322, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0036466213583672413, "chaser_grad_norm": 0.5243135318160057, "chaser_new_fields": 54.22500228881836, "chaser_partner_vision": 0.9665000438690186, "chaser_return": 18.047500610351562, "chaser_subspace_norm": 224.19346618652344, "collisions_per_episode": 16.42500114440918, "entropy": 0.19543031975626945, "explorer_grad_norm": 0.7758239656686783, "explorer_new_fields": 60.57500076293945, "explorer_partner_vision": 0.9665000438690186, "explorer_return": 27.887500762939453, "explorer_subspace_norm": 80.1183853149414, "final_distance": 1.592462182044983, "max_param_abs": 2.518620252609253, "policy_loss": -0.00373345779814116, "value_loss": 3.3543198704719543}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:35:20.165250+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020920.165264, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.192060884389696, "elapsed_seconds": 18691.419183802907, "env_steps_per_second": 4066.036893862932, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27326636233629986, "eta_seconds": 983.7589044106794, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0025184930163502095, "chaser_grad_norm": 0.6625885665416718, "chaser_new_fields": 63.42499923706055, "chaser_partner_vision": 0.9562500715255737, "chaser_return": 12.032500267028809, "chaser_subspace_norm": 240.88145446777344, "collisions_per_episode": 9.02500057220459, "entropy": 0.1566174291074276, "explorer_grad_norm": 1.0039133876562119, "explorer_new_fields": 69.80000305175781, "explorer_partner_vision": 0.9562500715255737, "explorer_return": 46.85000228881836, "explorer_subspace_norm": 91.64797973632812, "final_distance": 1.5585548877716064, "max_param_abs": 2.4977643489837646, "policy_loss": -0.006163327036418309, "value_loss": 2.630453050136566}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:43:23.367243+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021403.3672569, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.329666914930826, "elapsed_seconds": 19186.800893750973, "env_steps_per_second": 4065.2947008692904, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13665812602386734, "eta_seconds": 491.9692536859224, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0037499017243796118, "chaser_grad_norm": 0.6010967083275318, "chaser_new_fields": 59.42499923706055, "chaser_partner_vision": 0.9705000519752502, "chaser_return": 13.805000305175781, "chaser_subspace_norm": 240.27850341796875, "collisions_per_episode": 11.300000190734863, "entropy": 0.16362516954541206, "explorer_grad_norm": 0.7001258432865143, "explorer_new_fields": 66.17500305175781, "explorer_partner_vision": 0.9705000519752502, "explorer_return": 40.650001525878906, "explorer_subspace_norm": 90.2586441040039, "final_distance": 1.5025851726531982, "max_param_abs": 2.494504451751709, "policy_loss": -0.0031281966454912435, "value_loss": 3.4149429202079773}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:51:38.748955+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021898.7489693, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.463953162627198, "elapsed_seconds": 19670.23138545791, "env_steps_per_second": 4067.0594276355864, "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-transformer/social/seed_0000/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:59:42.179446+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022382.17946, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.463973339191095, "elapsed_seconds": 19670.30402108794, "env_steps_per_second": 4067.04440939166, "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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0030988860332676893, "chaser_grad_norm": 0.36728256940841675, "chaser_new_fields": 62.0, "chaser_partner_vision": 0.9725000262260437, "chaser_return": 13.717500686645508, "chaser_subspace_norm": 272.79351806640625, "collisions_per_episode": 10.925000190734863, "entropy": 0.11317355558276176, "explorer_grad_norm": 0.8740078210830688, "explorer_new_fields": 69.95000457763672, "explorer_partner_vision": 0.9725000262260437, "explorer_return": 46.98749923706055, "explorer_subspace_norm": 102.8924560546875, "final_distance": 1.6796330213546753, "max_param_abs": 2.4855704307556152, "policy_loss": -0.0027665531197555993, "value_loss": 1.877354085445404}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:59:42.252081+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022382.2520971, "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..101b2b68ee2f18adeed24cbcbf9512717b156872 --- /dev/null +++ b/social/seed_0000/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.463973339191095, + "elapsed_seconds": 19670.30402108794, + "env_steps_per_second": 4067.04440939166, + "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-transformer/social/seed_0000/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0030988860332676893, + "chaser_grad_norm": 0.36728256940841675, + "chaser_new_fields": 62.0, + "chaser_partner_vision": 0.9725000262260437, + "chaser_return": 13.717500686645508, + "chaser_subspace_norm": 272.79351806640625, + "collisions_per_episode": 10.925000190734863, + "entropy": 0.11317355558276176, + "explorer_grad_norm": 0.8740078210830688, + "explorer_new_fields": 69.95000457763672, + "explorer_partner_vision": 0.9725000262260437, + "explorer_return": 46.98749923706055, + "explorer_subspace_norm": 102.8924560546875, + "final_distance": 1.6796330213546753, + "max_param_abs": 2.4855704307556152, + "policy_loss": -0.0027665531197555993, + "value_loss": 1.877354085445404 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:59:42.252081+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022382.2520971, + "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..eead7ba441fd06374ff88180a55b94d0d5137ba1 --- /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:f65e21859c2db411274fb48eff4489b878589dc01d994df540c2fd45a29d7070 +size 29760384 diff --git a/social/seed_0001/attempt_01/config.json b/social/seed_0001/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..57509e78ce6ff61b6f3eced8a0961e5575b51808 --- /dev/null +++ b/social/seed_0001/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0001/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..01706d04f1476c9509ffca960a89e3ac8189dd1b --- /dev/null +++ b/social/seed_0001/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007143604480532101, "elapsed_seconds": 2.5716976129915565, "env_steps_per_second": 1555.392819044131, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.28649460061615, "eta_seconds": 51431.38056221814, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.006448477307564815, "chaser_grad_norm": 2.3706833124160767, "chaser_new_fields": 34.45000076293945, "chaser_partner_vision": 0.3450000286102295, "chaser_return": -1.3050004243850708, "chaser_subspace_norm": 0.0396004281938076, "collisions_per_episode": 1.7000000476837158, "entropy": 2.7463639974594116, "explorer_grad_norm": 8.532778143882751, "explorer_new_fields": 35.625, "explorer_partner_vision": 0.3450000286102295, "explorer_return": 2.2125000953674316, "explorer_subspace_norm": 0.03002295270562172, "final_distance": 4.862380504608154, "max_param_abs": 1.0011910200119019, "policy_loss": -0.01340884896740313, "value_loss": 6.868173539638519}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.625788+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.625805, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13423186734250825, "elapsed_seconds": 483.2347224330297, "env_steps_per_second": 4138.775438011235, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.235042826357822, "eta_seconds": 18846.15417488816, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004304458694286861, "chaser_grad_norm": 0.6996965259313583, "chaser_new_fields": 53.25, "chaser_partner_vision": 0.5190000534057617, "chaser_return": 16.817501068115234, "chaser_subspace_norm": 9.939868927001953, "collisions_per_episode": 14.975000381469727, "entropy": 0.7790972888469696, "explorer_grad_norm": 2.0114499628543854, "explorer_new_fields": 58.45000076293945, "explorer_partner_vision": 0.5190000534057617, "explorer_return": 27.600000381469727, "explorer_subspace_norm": 4.980358600616455, "final_distance": 4.814571857452393, "max_param_abs": 1.2308900356292725, "policy_loss": -0.005673073314135024, "value_loss": 4.247539281845093}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:39:55.288812+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003195.2888231, "update": 500} +{"device": "cuda", "elapsed_hours": 0.26855326387555883, "elapsed_seconds": 966.7917499520117, "env_steps_per_second": 4137.395669955341, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.102512013635618, "eta_seconds": 18369.043249088223, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0036183392321760266, "chaser_grad_norm": 1.0195853263139725, "chaser_new_fields": 54.20000076293945, "chaser_partner_vision": 0.5315000414848328, "chaser_return": 11.767500877380371, "chaser_subspace_norm": 18.132001876831055, "collisions_per_episode": 10.175000190734863, "entropy": 0.4758073091506958, "explorer_grad_norm": 2.03205868601799, "explorer_new_fields": 65.55000305175781, "explorer_partner_vision": 0.5315000414848328, "explorer_return": 40.57500076293945, "explorer_subspace_norm": 12.300207138061523, "final_distance": 4.186269760131836, "max_param_abs": 1.37955904006958, "policy_loss": -0.005668057265694415, "value_loss": 3.295992136001587}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:47:58.845839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003678.845853, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4019966364516747, "elapsed_seconds": 1447.187891226029, "env_steps_per_second": 4145.971671250592, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.957958516237321, "eta_seconds": 17848.650658454357, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.003112610729185228, "chaser_grad_norm": 0.8551076203584671, "chaser_new_fields": 53.79999923706055, "chaser_partner_vision": 0.5697500109672546, "chaser_return": 10.170001029968262, "chaser_subspace_norm": 22.825389862060547, "collisions_per_episode": 8.90000057220459, "entropy": 0.461285725235939, "explorer_grad_norm": 2.000103861093521, "explorer_new_fields": 67.45000457763672, "explorer_partner_vision": 0.5697500109672546, "explorer_return": 43.5, "explorer_subspace_norm": 15.428488731384277, "final_distance": 3.2252197265625, "max_param_abs": 1.4746465682983398, "policy_loss": -0.007614844674247223, "value_loss": 2.9343146681785583}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:55:59.241978+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004159.24199, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5371594234391684, "elapsed_seconds": 1933.7739243810065, "env_steps_per_second": 4136.988248282833, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.834434810952517, "eta_seconds": 17403.96531942906, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.006146629568725359, "chaser_grad_norm": 1.126807078719139, "chaser_new_fields": 58.375, "chaser_partner_vision": 0.4982500374317169, "chaser_return": 8.285000801086426, "chaser_subspace_norm": 32.49472427368164, "collisions_per_episode": 6.200000286102295, "entropy": 0.3478881046175957, "explorer_grad_norm": 1.6097354590892792, "explorer_new_fields": 68.0999984741211, "explorer_partner_vision": 0.4982500374317169, "explorer_return": 47.212501525878906, "explorer_subspace_norm": 22.67452049255371, "final_distance": 3.3293514251708984, "max_param_abs": 1.6135332584381104, "policy_loss": -0.008130347678431349, "value_loss": 2.7667242884635925}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:05.828013+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004645.828035, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6726283286536151, "elapsed_seconds": 2421.4619831530144, "env_steps_per_second": 4129.736526765075, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.708398300575306, "eta_seconds": 16950.2338820711, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.004809704232840772, "chaser_grad_norm": 0.6610389649868011, "chaser_new_fields": 60.70000076293945, "chaser_partner_vision": 0.48075002431869507, "chaser_return": 9.447501182556152, "chaser_subspace_norm": 32.11764907836914, "collisions_per_episode": 6.875, "entropy": 0.3576778471469879, "explorer_grad_norm": 1.4627466201782227, "explorer_new_fields": 70.875, "explorer_partner_vision": 0.48075002431869507, "explorer_return": 50.587501525878906, "explorer_subspace_norm": 27.556615829467773, "final_distance": 3.718161106109619, "max_param_abs": 1.7238563299179077, "policy_loss": -0.00884396871738602, "value_loss": 2.8377037048339844}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:13.516071+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005133.5160825, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8073911946389125, "elapsed_seconds": 2906.6083007000852, "env_steps_per_second": 4128.523267861611, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.575216769620504, "eta_seconds": 16470.780370633816, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004670768160804251, "chaser_grad_norm": 0.6260906308889389, "chaser_new_fields": 55.10000228881836, "chaser_partner_vision": 0.6022500395774841, "chaser_return": 11.205000877380371, "chaser_subspace_norm": 38.28797149658203, "collisions_per_episode": 9.5, "entropy": 0.34857507050037384, "explorer_grad_norm": 1.76337131857872, "explorer_new_fields": 65.6500015258789, "explorer_partner_vision": 0.6022500395774841, "explorer_return": 40.5, "explorer_subspace_norm": 31.415590286254883, "final_distance": 2.779783248901367, "max_param_abs": 1.7717581987380981, "policy_loss": -0.008363365421999447, "value_loss": 3.3369017839431763}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:18.662389+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005618.6624026, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.944358372294454, "elapsed_seconds": 3399.6901402600342, "env_steps_per_second": 4118.022355687149, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.451975183673855, "eta_seconds": 16027.110661225877, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.003644548550951754, "chaser_grad_norm": 1.0064982920885086, "chaser_new_fields": 49.32500076293945, "chaser_partner_vision": 0.8072500228881836, "chaser_return": 14.680000305175781, "chaser_subspace_norm": 44.26333999633789, "collisions_per_episode": 13.949999809265137, "entropy": 0.34912969172000885, "explorer_grad_norm": 2.093039572238922, "explorer_new_fields": 66.0999984741211, "explorer_partner_vision": 0.8072500228881836, "explorer_return": 37.07500076293945, "explorer_subspace_norm": 30.736562728881836, "final_distance": 3.216071844100952, "max_param_abs": 1.808855652809143, "policy_loss": -0.008007420041593605, "value_loss": 3.789446771144867}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:31.744229+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006111.74424, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0825982640588903, "elapsed_seconds": 3897.3537506120047, "env_steps_per_second": 4105.349686947844, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.330393056235561, "eta_seconds": 15589.415002448019, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.002988333394156939, "chaser_grad_norm": 1.6485087871551514, "chaser_new_fields": 48.92499923706055, "chaser_partner_vision": 0.8860000371932983, "chaser_return": 22.94500160217285, "chaser_subspace_norm": 60.09316635131836, "collisions_per_episode": 21.600000381469727, "entropy": 0.29308756440877914, "explorer_grad_norm": 1.802350491285324, "explorer_new_fields": 60.900001525878906, "explorer_partner_vision": 0.8860000371932983, "explorer_return": 25.524999618530273, "explorer_subspace_norm": 32.74444580078125, "final_distance": 2.215254306793213, "max_param_abs": 1.879026174545288, "policy_loss": -0.003483624539811103, "value_loss": 4.468444466590881}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:36:49.407839+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006609.4078538, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.217639241822245, "elapsed_seconds": 4383.501270560082, "env_steps_per_second": 4106.306554737266, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.194090721832177, "eta_seconds": 15098.726598595838, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.003488023554446329, "chaser_grad_norm": 0.8570198118686676, "chaser_new_fields": 50.82500076293945, "chaser_partner_vision": 0.9277500510215759, "chaser_return": 19.582502365112305, "chaser_subspace_norm": 60.19320297241211, "collisions_per_episode": 18.225000381469727, "entropy": 0.3392892926931381, "explorer_grad_norm": 1.2249049246311188, "explorer_new_fields": 59.10000228881836, "explorer_partner_vision": 0.9277500510215759, "explorer_return": 23.575000762939453, "explorer_subspace_norm": 33.20096969604492, "final_distance": 1.9275315999984741, "max_param_abs": 1.922048568725586, "policy_loss": -0.008928013686102076, "value_loss": 3.7924684882164}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:44:55.555358+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007095.5553703, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3528149201950226, "elapsed_seconds": 4870.1337127020815, "env_steps_per_second": 4106.6634264757095, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.058444760585068, "eta_seconds": 14610.401138106246, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.006669415999672523, "chaser_grad_norm": 0.8532964140176773, "chaser_new_fields": 51.70000076293945, "chaser_partner_vision": 0.9545000195503235, "chaser_return": 19.01500129699707, "chaser_subspace_norm": 74.27543640136719, "collisions_per_episode": 17.600000381469727, "entropy": 0.2675098106265068, "explorer_grad_norm": 1.5492980480194092, "explorer_new_fields": 61.150001525878906, "explorer_partner_vision": 0.9545000195503235, "explorer_return": 27.30000114440918, "explorer_subspace_norm": 37.16187286376953, "final_distance": 1.8200578689575195, "max_param_abs": 1.9390627145767212, "policy_loss": -0.008118558697390199, "value_loss": 3.5549176931381226}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:02.187803+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007582.187819, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4882018217108433, "elapsed_seconds": 5357.526558159036, "env_steps_per_second": 4106.372551060145, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9234411663285873, "eta_seconds": 14124.388198782914, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0034825982061477134, "chaser_grad_norm": 1.3704558610916138, "chaser_new_fields": 44.625, "chaser_partner_vision": 0.9687500596046448, "chaser_return": 22.192502975463867, "chaser_subspace_norm": 63.529945373535156, "collisions_per_episode": 21.625, "entropy": 0.4013589695096016, "explorer_grad_norm": 1.2256624698638916, "explorer_new_fields": 49.22500228881836, "explorer_partner_vision": 0.9687500596046448, "explorer_return": 9.350000381469727, "explorer_subspace_norm": 28.638315200805664, "final_distance": 1.768484115600586, "max_param_abs": 1.9650145769119263, "policy_loss": -0.0035320273484096276, "value_loss": 5.443762421607971}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:09.580648+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008069.5806625, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6247733388614143, "elapsed_seconds": 5849.184019901091, "env_steps_per_second": 4103.136423532429, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.791137790676633, "eta_seconds": 13648.09604643588, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.005866068563413762, "chaser_grad_norm": 0.75855852663517, "chaser_new_fields": 49.25, "chaser_partner_vision": 0.9620000720024109, "chaser_return": 19.3700008392334, "chaser_subspace_norm": 90.9434814453125, "collisions_per_episode": 18.55000114440918, "entropy": 0.19908056780695915, "explorer_grad_norm": 2.1868440359830856, "explorer_new_fields": 58.900001525878906, "explorer_partner_vision": 0.9620000720024109, "explorer_return": 22.8125, "explorer_subspace_norm": 43.30324172973633, "final_distance": 1.6130472421646118, "max_param_abs": 2.00337290763855, "policy_loss": -0.003826605038719677, "value_loss": 3.905464768409729}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:21.238108+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008561.238122, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.761861779509733, "elapsed_seconds": 6342.702406235039, "env_steps_per_second": 4099.199100755749, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.659251388212523, "eta_seconds": 13173.304997565083, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.004516099351901559, "chaser_grad_norm": 0.9679954349994659, "chaser_new_fields": 49.0, "chaser_partner_vision": 0.955500066280365, "chaser_return": 18.172500610351562, "chaser_subspace_norm": 90.96558380126953, "collisions_per_episode": 17.325000762939453, "entropy": 0.21279576420783997, "explorer_grad_norm": 2.9747171103954315, "explorer_new_fields": 56.20000076293945, "explorer_partner_vision": 0.955500066280365, "explorer_return": 21.17500114440918, "explorer_subspace_norm": 46.61947250366211, "final_distance": 1.4821068048477173, "max_param_abs": 2.027212619781494, "policy_loss": -0.005418258045779112, "value_loss": 4.011760950088501}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:34.756493+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009054.7565055, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.9006082305594463, "elapsed_seconds": 6842.189630014007, "env_steps_per_second": 4092.257232564114, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5297009996104003, "eta_seconds": 12706.923598597441, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.003727432425300492, "chaser_grad_norm": 0.702857069671154, "chaser_new_fields": 53.97500228881836, "chaser_partner_vision": 0.9630000591278076, "chaser_return": 16.940000534057617, "chaser_subspace_norm": 102.85576629638672, "collisions_per_episode": 15.300000190734863, "entropy": 0.21533358097076416, "explorer_grad_norm": 1.308296263217926, "explorer_new_fields": 62.79999923706055, "explorer_partner_vision": 0.9630000591278076, "explorer_return": 31.037500381469727, "explorer_subspace_norm": 44.47128677368164, "final_distance": 1.5872725248336792, "max_param_abs": 2.0336663722991943, "policy_loss": -0.005088015817108271, "value_loss": 3.7025079131126404}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:54.243720+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009554.2437346, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0371871466527876, "elapsed_seconds": 7333.873727950035, "env_steps_per_second": 4090.6076533152423, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3953119110879793, "eta_seconds": 12223.122879916726, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004820447962025298, "chaser_grad_norm": 0.4992205277085304, "chaser_new_fields": 54.22500228881836, "chaser_partner_vision": 0.9655000567436218, "chaser_return": 15.977500915527344, "chaser_subspace_norm": 112.73866271972656, "collisions_per_episode": 14.425000190734863, "entropy": 0.2060987651348114, "explorer_grad_norm": 1.0820595473051071, "explorer_new_fields": 59.79999923706055, "explorer_partner_vision": 0.9655000567436218, "explorer_return": 27.649999618530273, "explorer_subspace_norm": 41.756309509277344, "final_distance": 1.6733932495117188, "max_param_abs": 2.0393853187561035, "policy_loss": -0.0047195199425260626, "value_loss": 3.211985230445862}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:34:05.927816+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010045.9278286, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1722059004044665, "elapsed_seconds": 7819.941241456079, "env_steps_per_second": 4092.102358820483, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2583088506066997, "eta_seconds": 11729.911862184119, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004632526282730254, "chaser_grad_norm": 1.0855955928564072, "chaser_new_fields": 47.35000228881836, "chaser_partner_vision": 0.9587500691413879, "chaser_return": 16.560001373291016, "chaser_subspace_norm": 116.50370788574219, "collisions_per_episode": 16.05000114440918, "entropy": 0.180303905159235, "explorer_grad_norm": 1.379273772239685, "explorer_new_fields": 51.625, "explorer_partner_vision": 0.9587500691413879, "explorer_return": 15.100000381469727, "explorer_subspace_norm": 46.3656005859375, "final_distance": 1.850345253944397, "max_param_abs": 2.061431646347046, "policy_loss": -0.0044207527339494845, "value_loss": 3.9910995960235596}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:11.995329+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010531.9953437, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.309602203611123, "elapsed_seconds": 8314.567933000042, "env_steps_per_second": 4089.2082756406326, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1247559225326955, "eta_seconds": 11249.121321117704, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.003587742456296894, "chaser_grad_norm": 0.5386281237006187, "chaser_new_fields": 53.525001525878906, "chaser_partner_vision": 0.9510000348091125, "chaser_return": 16.202499389648438, "chaser_subspace_norm": 131.96226501464844, "collisions_per_episode": 14.824999809265137, "entropy": 0.20087382569909096, "explorer_grad_norm": 1.1583024710416794, "explorer_new_fields": 61.5, "explorer_partner_vision": 0.9510000348091125, "explorer_return": 29.587499618530273, "explorer_subspace_norm": 46.46718978881836, "final_distance": 1.7569717168807983, "max_param_abs": 2.0682308673858643, "policy_loss": -0.007491338379665535, "value_loss": 3.5119765400886536}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:26.622022+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011026.622034, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4465418338455818, "elapsed_seconds": 8807.550601844094, "env_steps_per_second": 4087.4020062357004, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9902177969223773, "eta_seconds": 10764.784068920559, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.002751322695903724, "chaser_grad_norm": 1.0448336750268936, "chaser_new_fields": 55.85000228881836, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 13.0600004196167, "chaser_subspace_norm": 138.04742431640625, "collisions_per_episode": 11.25, "entropy": 0.21180135756731033, "explorer_grad_norm": 2.0704070031642914, "explorer_new_fields": 61.775001525878906, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 33.587501525878906, "explorer_subspace_norm": 46.88496780395508, "final_distance": 1.8486700057983398, "max_param_abs": 2.0947110652923584, "policy_loss": -0.005396676776001641, "value_loss": 3.6277953386306763}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:39.604690+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011519.6047053, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5846985578805715, "elapsed_seconds": 9304.914808370057, "env_steps_per_second": 4083.863289733489, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8567720902890525, "eta_seconds": 10284.37952504059, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0038068282922196772, "chaser_grad_norm": 0.4306594729423523, "chaser_new_fields": 56.57500076293945, "chaser_partner_vision": 0.9670000672340393, "chaser_return": 14.3774995803833, "chaser_subspace_norm": 134.6617889404297, "collisions_per_episode": 12.574999809265137, "entropy": 0.18146897107362747, "explorer_grad_norm": 1.542133778333664, "explorer_new_fields": 62.85000228881836, "explorer_partner_vision": 0.9670000672340393, "explorer_return": 33.48749923706055, "explorer_subspace_norm": 51.04337692260742, "final_distance": 1.8353703022003174, "max_param_abs": 2.1258153915405273, "policy_loss": -0.005330101021570499, "value_loss": 3.0833284854888916}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:56.968895+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012016.968907, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7196183983105584, "elapsed_seconds": 9790.62623391801, "env_steps_per_second": 4085.5405000986143, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7196183983105584, "eta_seconds": 9790.62623391801, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.003712276042431739, "chaser_grad_norm": 0.48314884305000305, "chaser_new_fields": 54.20000076293945, "chaser_partner_vision": 0.9675000309944153, "chaser_return": 13.847500801086426, "chaser_subspace_norm": 137.04202270507812, "collisions_per_episode": 12.375, "entropy": 0.1987452618777752, "explorer_grad_norm": 1.6377961188554764, "explorer_new_fields": 63.900001525878906, "explorer_partner_vision": 0.9675000309944153, "explorer_return": 34.86249923706055, "explorer_subspace_norm": 54.75956726074219, "final_distance": 1.6417652368545532, "max_param_abs": 2.1501688957214355, "policy_loss": -0.0055382894408069205, "value_loss": 2.9678056240081787}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:15:02.680323+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012502.680337, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.856108171771953, "elapsed_seconds": 10281.989418379031, "env_steps_per_second": 4084.8126068798615, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5840978696984336, "eta_seconds": 9302.752330914362, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0041013707123234155, "chaser_grad_norm": 0.5680395811796188, "chaser_new_fields": 57.20000076293945, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 15.397500991821289, "chaser_subspace_norm": 152.85302734375, "collisions_per_episode": 13.425000190734863, "entropy": 0.17234127968549728, "explorer_grad_norm": 1.3457107841968536, "explorer_new_fields": 66.7750015258789, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 37.86249923706055, "explorer_subspace_norm": 55.45365524291992, "final_distance": 1.5485280752182007, "max_param_abs": 2.2032127380371094, "policy_loss": -0.004376603524207212, "value_loss": 2.7350229620933533}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:14.043508+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012994.0435212, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9932733193125265, "elapsed_seconds": 10775.783949525096, "env_steps_per_second": 4083.2296013079535, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4490418067102486, "eta_seconds": 8816.550504156896, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0033466992302075838, "chaser_grad_norm": 0.7450225502252579, "chaser_new_fields": 54.17499923706055, "chaser_partner_vision": 0.9612500667572021, "chaser_return": 15.857498168945312, "chaser_subspace_norm": 155.0267791748047, "collisions_per_episode": 14.324999809265137, "entropy": 0.18501728028059006, "explorer_grad_norm": 0.9732646644115448, "explorer_new_fields": 62.97500228881836, "explorer_partner_vision": 0.9612500667572021, "explorer_return": 32.42499923706055, "explorer_subspace_norm": 62.69338607788086, "final_distance": 1.6271203756332397, "max_param_abs": 2.2014667987823486, "policy_loss": -0.0064696257877585595, "value_loss": 3.246390402317047}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:27.838038+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013487.8380516, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1272685478677804, "elapsed_seconds": 11258.16677232401, "env_steps_per_second": 4085.9227732424392, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.31145936146749, "eta_seconds": 8321.253701282963, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.003819246445424085, "chaser_grad_norm": 0.6606405824422836, "chaser_new_fields": 59.54999923706055, "chaser_partner_vision": 0.9747500419616699, "chaser_return": 13.727499961853027, "chaser_subspace_norm": 171.49069213867188, "collisions_per_episode": 11.375, "entropy": 0.15117892250418663, "explorer_grad_norm": 0.9140024036169052, "explorer_new_fields": 65.3499984741211, "explorer_partner_vision": 0.9747500419616699, "explorer_return": 38.32500076293945, "explorer_subspace_norm": 61.371307373046875, "final_distance": 1.5088834762573242, "max_param_abs": 2.2150259017944336, "policy_loss": -0.005682039295315633, "value_loss": 2.377293050289154}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:30.220860+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013970.220872, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2614251002947436, "elapsed_seconds": 11741.130361061078, "env_steps_per_second": 4088.1924077080184, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1742834001964955, "eta_seconds": 7827.420240707384, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.002100970029541571, "chaser_grad_norm": 0.6557553559541702, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9707500338554382, "chaser_return": 11.895001411437988, "chaser_subspace_norm": 160.04888916015625, "collisions_per_episode": 10.050000190734863, "entropy": 0.15994855761528015, "explorer_grad_norm": 1.1096415966749191, "explorer_new_fields": 64.5250015258789, "explorer_partner_vision": 0.9707500338554382, "explorer_return": 37.45000076293945, "explorer_subspace_norm": 65.14669036865234, "final_distance": 1.8222700357437134, "max_param_abs": 2.2230629920959473, "policy_loss": -0.004625786159181189, "value_loss": 2.558869481086731}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:33.184450+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014453.1844642, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.396083948333916, "elapsed_seconds": 12225.902214002097, "env_steps_per_second": 4089.6777288743515, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0376503690003496, "eta_seconds": 7335.541328401258, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0027315824771960706, "chaser_grad_norm": 0.6480161547660828, "chaser_new_fields": 56.07500076293945, "chaser_partner_vision": 0.9667500257492065, "chaser_return": 15.214999198913574, "chaser_subspace_norm": 148.00718688964844, "collisions_per_episode": 13.40000057220459, "entropy": 0.2123485654592514, "explorer_grad_norm": 0.9570952206850052, "explorer_new_fields": 63.17499923706055, "explorer_partner_vision": 0.9667500257492065, "explorer_return": 33.63750076293945, "explorer_subspace_norm": 63.0347900390625, "final_distance": 1.7802537679672241, "max_param_abs": 2.25895619392395, "policy_loss": -0.005573465166612834, "value_loss": 3.0373697876930237}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:37.956301+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014937.9563127, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5344926806066845, "elapsed_seconds": 12724.173650184064, "env_steps_per_second": 4086.7093950142516, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9031883664805223, "eta_seconds": 6851.47811932988, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.002661498861193934, "chaser_grad_norm": 0.45983535796403885, "chaser_new_fields": 59.70000076293945, "chaser_partner_vision": 0.9617500305175781, "chaser_return": 12.635001182556152, "chaser_subspace_norm": 185.69851684570312, "collisions_per_episode": 10.350000381469727, "entropy": 0.1623547337949276, "explorer_grad_norm": 1.7783105373382568, "explorer_new_fields": 67.2249984741211, "explorer_partner_vision": 0.9617500305175781, "explorer_return": 41.650001525878906, "explorer_subspace_norm": 68.91088104248047, "final_distance": 1.546286702156067, "max_param_abs": 2.296994209289551, "policy_loss": -0.004194682253483961, "value_loss": 2.6339756846427917}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:56.227740+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015436.2277558, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6706047893344658, "elapsed_seconds": 13214.177241604077, "env_steps_per_second": 4086.5200316811333, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7673282319017798, "eta_seconds": 6362.381634846407, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.002595398165151863, "chaser_grad_norm": 0.4217633381485939, "chaser_new_fields": 51.72500228881836, "chaser_partner_vision": 0.9702500700950623, "chaser_return": 13.644999504089355, "chaser_subspace_norm": 165.349365234375, "collisions_per_episode": 12.65000057220459, "entropy": 0.1603407934308052, "explorer_grad_norm": 0.8940415382385254, "explorer_new_fields": 58.42499923706055, "explorer_partner_vision": 0.9702500700950623, "explorer_return": 27.9375, "explorer_subspace_norm": 78.28569030761719, "final_distance": 1.6670655012130737, "max_param_abs": 2.360933780670166, "policy_loss": -0.003905867955637632, "value_loss": 3.424300730228424}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:12:06.231328+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015926.2313406, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.8087351186538903, "elapsed_seconds": 13711.446427154006, "env_steps_per_second": 4084.178886415527, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6323150508516673, "eta_seconds": 5876.334183066002, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.003417840536535266, "chaser_grad_norm": 0.5128882080316544, "chaser_new_fields": 53.75, "chaser_partner_vision": 0.9547500610351562, "chaser_return": 16.142499923706055, "chaser_subspace_norm": 167.82554626464844, "collisions_per_episode": 14.725000381469727, "entropy": 0.1955343410372734, "explorer_grad_norm": 1.0539238452911377, "explorer_new_fields": 61.75, "explorer_partner_vision": 0.9547500610351562, "explorer_return": 30.162500381469727, "explorer_subspace_norm": 82.08577728271484, "final_distance": 1.6971056461334229, "max_param_abs": 2.424859046936035, "policy_loss": -0.005180578025061777, "value_loss": 3.1520726084709167}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:23.500516+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016423.5005317, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9426730444789024, "elapsed_seconds": 14193.622960124048, "env_steps_per_second": 4086.342166686179, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4954966720437215, "eta_seconds": 5383.788019357397, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.006948581867550985, "chaser_grad_norm": 1.5984272360801697, "chaser_new_fields": 53.25, "chaser_partner_vision": 0.9720000624656677, "chaser_return": 15.25, "chaser_subspace_norm": 165.09585571289062, "collisions_per_episode": 13.90000057220459, "entropy": 0.21762651950120926, "explorer_grad_norm": 0.9787304550409317, "explorer_new_fields": 60.275001525878906, "explorer_partner_vision": 0.9720000624656677, "explorer_return": 28.587499618530273, "explorer_subspace_norm": 73.781982421875, "final_distance": 1.9751795530319214, "max_param_abs": 2.459953546524048, "policy_loss": -0.005152379174717225, "value_loss": 4.135168671607971}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:25.677047+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016905.6770594, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.079638605082226, "elapsed_seconds": 14686.698978296015, "env_steps_per_second": 4085.3291872236177, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3598795350274089, "eta_seconds": 4895.566326098672, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0024383773270830744, "chaser_grad_norm": 0.7703244090080261, "chaser_new_fields": 56.400001525878906, "chaser_partner_vision": 0.9687500596046448, "chaser_return": 15.655001640319824, "chaser_subspace_norm": 193.86915588378906, "collisions_per_episode": 13.850000381469727, "entropy": 0.16178318485617638, "explorer_grad_norm": 1.408334419131279, "explorer_new_fields": 62.35000228881836, "explorer_partner_vision": 0.9687500596046448, "explorer_return": 31.61250114440918, "explorer_subspace_norm": 87.53654479980469, "final_distance": 1.5938421487808228, "max_param_abs": 2.5002524852752686, "policy_loss": -0.004136937701133547, "value_loss": 3.157750904560089}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:38.753067+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017398.7530825, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.215398495231671, "elapsed_seconds": 15175.434582834016, "env_steps_per_second": 4085.5502135097004, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2238253695833883, "eta_seconds": 4405.771330500198, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00395154064258918, "chaser_grad_norm": 0.49677933007478714, "chaser_new_fields": 53.22500228881836, "chaser_partner_vision": 0.9672500491142273, "chaser_return": 11.317500114440918, "chaser_subspace_norm": 210.26014709472656, "collisions_per_episode": 10.175000190734863, "entropy": 0.20660146325826645, "explorer_grad_norm": 1.0447909981012344, "explorer_new_fields": 59.67499923706055, "explorer_partner_vision": 0.9672500491142273, "explorer_return": 31.012500762939453, "explorer_subspace_norm": 73.98773193359375, "final_distance": 1.7164363861083984, "max_param_abs": 2.5166351795196533, "policy_loss": -0.005759350895857551, "value_loss": 3.2624166011810303}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:47.488671+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017887.4886825, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.352478085823887, "elapsed_seconds": 15668.921108965995, "env_steps_per_second": 4084.518618411974, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0881195214559718, "eta_seconds": 3917.2302772414987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0022466290782698906, "chaser_grad_norm": 0.9437057971954346, "chaser_new_fields": 60.17499923706055, "chaser_partner_vision": 0.9732500314712524, "chaser_return": 9.46250057220459, "chaser_subspace_norm": 195.47817993164062, "collisions_per_episode": 7.175000190734863, "entropy": 0.17338208481669426, "explorer_grad_norm": 0.7868348658084869, "explorer_new_fields": 68.4000015258789, "explorer_partner_vision": 0.9732500314712524, "explorer_return": 46.3125, "explorer_subspace_norm": 77.03369903564453, "final_distance": 1.8358350992202759, "max_param_abs": 2.5263357162475586, "policy_loss": -0.006061806699264594, "value_loss": 2.71109938621521}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:53:00.975197+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018380.9752102, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.491627811768622, "elapsed_seconds": 16169.860122367041, "env_steps_per_second": 4081.6679612896073, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9527695358297078, "eta_seconds": 3429.970328986948, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0017677274315848202, "chaser_grad_norm": 0.9131149053573608, "chaser_new_fields": 59.150001525878906, "chaser_partner_vision": 0.9702500700950623, "chaser_return": 9.440000534057617, "chaser_subspace_norm": 200.4278106689453, "collisions_per_episode": 7.400000095367432, "entropy": 0.17324408888816833, "explorer_grad_norm": 1.060551330447197, "explorer_new_fields": 64.9749984741211, "explorer_partner_vision": 0.9702500700950623, "explorer_return": 41.73749923706055, "explorer_subspace_norm": 79.14791107177734, "final_distance": 1.5274208784103394, "max_param_abs": 2.544046401977539, "policy_loss": -0.005810583109052914, "value_loss": 2.6551509499549866}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:01:21.914209+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018881.9142208, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.626936523911395, "elapsed_seconds": 16656.97148608102, "env_steps_per_second": 4082.3747616319383, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8165182101020109, "eta_seconds": 2939.465556367239, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.003169204183514296, "chaser_grad_norm": 0.6116224676370621, "chaser_new_fields": 56.72500228881836, "chaser_partner_vision": 0.9675000309944153, "chaser_return": 12.77750015258789, "chaser_subspace_norm": 207.05908203125, "collisions_per_episode": 10.975000381469727, "entropy": 0.1828526295721531, "explorer_grad_norm": 1.2805046290159225, "explorer_new_fields": 63.92499923706055, "explorer_partner_vision": 0.9675000309944153, "explorer_return": 37.4375, "explorer_subspace_norm": 92.22782897949219, "final_distance": 1.595851182937622, "max_param_abs": 2.5995171070098877, "policy_loss": -0.0013983421665670903, "value_loss": 3.1321605443954468}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:09:29.025574+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019369.0255895, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.760834607915571, "elapsed_seconds": 17139.004588496056, "env_steps_per_second": 4084.251196652634, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6801192297022245, "eta_seconds": 2448.429226928008, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.004214868916721848, "chaser_grad_norm": 0.4559861272573471, "chaser_new_fields": 56.375, "chaser_partner_vision": 0.9387500286102295, "chaser_return": 12.352500915527344, "chaser_subspace_norm": 229.63870239257812, "collisions_per_episode": 10.725000381469727, "entropy": 0.15422671288251877, "explorer_grad_norm": 1.5533980131149292, "explorer_new_fields": 64.25, "explorer_partner_vision": 0.9387500286102295, "explorer_return": 35.95000076293945, "explorer_subspace_norm": 92.24241638183594, "final_distance": 1.5838836431503296, "max_param_abs": 2.67535662651062, "policy_loss": -0.004316889022959458, "value_loss": 2.8737707138061523}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:17:31.058676+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019851.0586882, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.894762310525015, "elapsed_seconds": 17621.14431789005, "env_steps_per_second": 4086.0002449955105, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5438624789472238, "eta_seconds": 1957.904924210006, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.007792032853078723, "chaser_grad_norm": 0.470505204051733, "chaser_new_fields": 59.10000228881836, "chaser_partner_vision": 0.9760000705718994, "chaser_return": 11.607501029968262, "chaser_subspace_norm": 229.90663146972656, "collisions_per_episode": 9.52500057220459, "entropy": 0.1427590511739254, "explorer_grad_norm": 1.1371123641729355, "explorer_new_fields": 65.42500305175781, "explorer_partner_vision": 0.9760000705718994, "explorer_return": 39.962501525878906, "explorer_subspace_norm": 93.0556640625, "final_distance": 1.5586512088775635, "max_param_abs": 2.7568931579589844, "policy_loss": -0.004322844569102813, "value_loss": 2.6408538222312927}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:33.198407+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020333.1984222, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.030274183728066, "elapsed_seconds": 18108.987061421038, "env_steps_per_second": 4086.368815053597, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4078600689509243, "eta_seconds": 1468.2962482233274, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.007198342805931901, "chaser_grad_norm": 0.5620444864034653, "chaser_new_fields": 48.5, "chaser_partner_vision": 0.9487500190734863, "chaser_return": 19.42500114440918, "chaser_subspace_norm": 249.6824188232422, "collisions_per_episode": 18.350000381469727, "entropy": 0.16947517544031143, "explorer_grad_norm": 1.1757971495389938, "explorer_new_fields": 53.20000076293945, "explorer_partner_vision": 0.9487500190734863, "explorer_return": 18.149999618530273, "explorer_subspace_norm": 97.8075180053711, "final_distance": 1.5503050088882446, "max_param_abs": 2.7804157733917236, "policy_loss": -0.0032951137279155773, "value_loss": 4.383156895637512}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:41.041149+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020821.0411613, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.1670552278366655, "elapsed_seconds": 18601.398820211994, "env_steps_per_second": 4085.7142376529, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27195027514929815, "eta_seconds": 979.0209905374734, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.00386411324489222, "chaser_grad_norm": 1.0835418552160263, "chaser_new_fields": 57.900001525878906, "chaser_partner_vision": 0.9692500233650208, "chaser_return": 12.102500915527344, "chaser_subspace_norm": 218.51898193359375, "collisions_per_episode": 10.074999809265137, "entropy": 0.16452598944306374, "explorer_grad_norm": 0.9866618514060974, "explorer_new_fields": 62.85000228881836, "explorer_partner_vision": 0.9692500233650208, "explorer_return": 36.76250076293945, "explorer_subspace_norm": 97.25251007080078, "final_distance": 1.702256441116333, "max_param_abs": 2.8538239002227783, "policy_loss": -0.0034050465477619696, "value_loss": 3.086686372756958}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:53.452909+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021313.452924, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.302075169382248, "elapsed_seconds": 19087.470609776094, "env_steps_per_second": 4086.450300023015, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1359506453687756, "eta_seconds": 489.42232332759215, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0018806820540220315, "chaser_grad_norm": 0.4889792241156101, "chaser_new_fields": 62.85000228881836, "chaser_partner_vision": 0.9790000319480896, "chaser_return": 11.420001029968262, "chaser_subspace_norm": 266.67926025390625, "collisions_per_episode": 8.550000190734863, "entropy": 0.1494312547147274, "explorer_grad_norm": 1.3342660516500473, "explorer_new_fields": 70.0250015258789, "explorer_partner_vision": 0.9790000319480896, "explorer_return": 47.125, "explorer_subspace_norm": 99.142578125, "final_distance": 1.577653169631958, "max_param_abs": 2.8910109996795654, "policy_loss": -0.0029456372307663514, "value_loss": 2.6880043745040894}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:59.524698+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021799.524711, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.435676280673619, "elapsed_seconds": 19568.43461042503, "env_steps_per_second": 4088.216640353042, "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-transformer/social/seed_0001/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:00.488699+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022280.488713, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.435685596511111, "elapsed_seconds": 19568.46814744, "env_steps_per_second": 4088.2096338473903, "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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0032518112735848703, "chaser_grad_norm": 0.4799686074256897, "chaser_new_fields": 58.82500076293945, "chaser_partner_vision": 0.9635000228881836, "chaser_return": 13.932499885559082, "chaser_subspace_norm": 267.9047546386719, "collisions_per_episode": 11.824999809265137, "entropy": 0.16130381077528, "explorer_grad_norm": 1.6371502578258514, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9635000228881836, "explorer_return": 35.51250076293945, "explorer_subspace_norm": 96.23020935058594, "final_distance": 1.6419010162353516, "max_param_abs": 2.923971652984619, "policy_loss": -0.0037321077841312444, "value_loss": 3.0047545433044434}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:58:00.522236+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022280.5222507, "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..9eb6af78f8cf5c32af3503a7b6f0a28e913983aa --- /dev/null +++ b/social/seed_0001/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.435685596511111, + "elapsed_seconds": 19568.46814744, + "env_steps_per_second": 4088.2096338473903, + "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-transformer/social/seed_0001/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0032518112735848703, + "chaser_grad_norm": 0.4799686074256897, + "chaser_new_fields": 58.82500076293945, + "chaser_partner_vision": 0.9635000228881836, + "chaser_return": 13.932499885559082, + "chaser_subspace_norm": 267.9047546386719, + "collisions_per_episode": 11.824999809265137, + "entropy": 0.16130381077528, + "explorer_grad_norm": 1.6371502578258514, + "explorer_new_fields": 63.85000228881836, + "explorer_partner_vision": 0.9635000228881836, + "explorer_return": 35.51250076293945, + "explorer_subspace_norm": 96.23020935058594, + "final_distance": 1.6419010162353516, + "max_param_abs": 2.923971652984619, + "policy_loss": -0.0037321077841312444, + "value_loss": 3.0047545433044434 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:58:00.522236+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022280.5222507, + "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..c96c74215ac73696695ddf7dccb2227e94b500cd --- /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:c1f4144d32c140593489ce5522c50515a469ce6c2ad5aae7e6125078757e8ada +size 29760368 diff --git a/social/seed_0002/attempt_01/config.json b/social/seed_0002/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..08b265d649c9273c005f5c87eb8994d3ce207f0e --- /dev/null +++ b/social/seed_0002/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0002/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..1e1406ab3b83a9a50bd9fbbb0271cbcf07c49b07 --- /dev/null +++ b/social/seed_0002/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007734034544490795, "elapsed_seconds": 2.7842524360166863, "env_steps_per_second": 1436.6513424776358, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 15.46729568552714, "eta_seconds": 55682.26446789771, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0025573000201966067, "chaser_grad_norm": 4.892130110412836, "chaser_new_fields": 33.32500076293945, "chaser_partner_vision": 0.3187500238418579, "chaser_return": -1.4650003910064697, "chaser_subspace_norm": 0.035075075924396515, "collisions_per_episode": 1.75, "entropy": 2.7578166127204895, "explorer_grad_norm": 6.5302894115448, "explorer_new_fields": 36.20000076293945, "explorer_partner_vision": 0.3187500238418579, "explorer_return": 3.049999952316284, "explorer_subspace_norm": 0.03145161643624306, "final_distance": 5.499385833740234, "max_param_abs": 1.0012017488479614, "policy_loss": -0.011323383716870694, "value_loss": 7.66939115524292}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.718976+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.7189908, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13558655094804836, "elapsed_seconds": 488.11158341297414, "env_steps_per_second": 4097.423761213776, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.287875486973887, "eta_seconds": 19036.35175310599, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005900153669618491, "chaser_grad_norm": 0.6113118529319763, "chaser_new_fields": 64.0250015258789, "chaser_partner_vision": 0.5015000104904175, "chaser_return": 12.202500343322754, "chaser_subspace_norm": 9.921438217163086, "collisions_per_episode": 8.77500057220459, "entropy": 0.439317524433136, "explorer_grad_norm": 1.8251222968101501, "explorer_new_fields": 73.45000457763672, "explorer_partner_vision": 0.5015000104904175, "explorer_return": 53.462501525878906, "explorer_subspace_norm": 10.071572303771973, "final_distance": 4.1822190284729, "max_param_abs": 1.2755125761032104, "policy_loss": -0.005785484785217676, "value_loss": 2.727020025253296}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:00.046306+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003200.0463178, "update": 500} +{"device": "cuda", "elapsed_hours": 0.273361418813066, "elapsed_seconds": 984.1011077270377, "env_steps_per_second": 4064.6230032590192, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.193866957448255, "eta_seconds": 18697.921046813717, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00445235120861831, "chaser_grad_norm": 1.0619208365678787, "chaser_new_fields": 64.67500305175781, "chaser_partner_vision": 0.5517500042915344, "chaser_return": 10.855000495910645, "chaser_subspace_norm": 17.411222457885742, "collisions_per_episode": 7.450000286102295, "entropy": 0.2593146935105324, "explorer_grad_norm": 1.3593606501817703, "explorer_new_fields": 78.5, "explorer_partner_vision": 0.5517500042915344, "explorer_return": 61.51250076293945, "explorer_subspace_norm": 21.687788009643555, "final_distance": 2.1962058544158936, "max_param_abs": 1.4078947305679321, "policy_loss": -0.006686117372481704, "value_loss": 2.4666616320610046}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:16.035830+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003696.035847, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4076381075163954, "elapsed_seconds": 1467.4971870590234, "env_steps_per_second": 4088.5938677841414, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.027536659368876, "eta_seconds": 18099.131973727955, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0037655055921586023, "chaser_grad_norm": 0.7338449358940125, "chaser_new_fields": 65.20000457763672, "chaser_partner_vision": 0.5555000305175781, "chaser_return": 11.407501220703125, "chaser_subspace_norm": 26.729408264160156, "collisions_per_episode": 7.825000286102295, "entropy": 0.23171206191182137, "explorer_grad_norm": 2.138176053762436, "explorer_new_fields": 76.80000305175781, "explorer_partner_vision": 0.5555000305175781, "explorer_return": 58.70000076293945, "explorer_subspace_norm": 25.008718490600586, "final_distance": 2.8716750144958496, "max_param_abs": 1.4692598581314087, "policy_loss": -0.00448155099293146, "value_loss": 2.3101357221603394}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:19.431908+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004179.4319189, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5440250548172237, "elapsed_seconds": 1958.4901973420056, "env_steps_per_second": 4084.779189018827, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.896225493355014, "eta_seconds": 17626.41177607805, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.003454054643468929, "chaser_grad_norm": 0.526061624288559, "chaser_new_fields": 67.6500015258789, "chaser_partner_vision": 0.5607500076293945, "chaser_return": 12.350001335144043, "chaser_subspace_norm": 33.780029296875, "collisions_per_episode": 8.300000190734863, "entropy": 0.16463036835193634, "explorer_grad_norm": 1.6877653896808624, "explorer_new_fields": 77.4000015258789, "explorer_partner_vision": 0.5607500076293945, "explorer_return": 59.13750076293945, "explorer_subspace_norm": 34.57163619995117, "final_distance": 2.0624797344207764, "max_param_abs": 1.505094051361084, "policy_loss": -0.0063155147097972986, "value_loss": 2.0022511184215546}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:30.424920+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004670.4249341, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6807064831227763, "elapsed_seconds": 2450.5433392419945, "env_steps_per_second": 4080.7276655197675, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.764945381859434, "eta_seconds": 17153.80337469396, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0057787601779279996, "chaser_grad_norm": 0.7621303349733353, "chaser_new_fields": 66.30000305175781, "chaser_partner_vision": 0.5675000548362732, "chaser_return": 14.625, "chaser_subspace_norm": 44.804779052734375, "collisions_per_episode": 10.75, "entropy": 0.13474775478243828, "explorer_grad_norm": 1.7482016682624817, "explorer_new_fields": 81.2750015258789, "explorer_partner_vision": 0.5675000548362732, "explorer_return": 61.625, "explorer_subspace_norm": 38.55632400512695, "final_distance": 2.124192476272583, "max_param_abs": 1.5256363153457642, "policy_loss": -0.0045058975154503855, "value_loss": 1.8514660894870758}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:42.478060+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005162.4780712, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8145359379122318, "elapsed_seconds": 2932.3293764840346, "env_steps_per_second": 4092.309716716892, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.615703648169314, "eta_seconds": 16616.533133409528, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.006340893610810028, "chaser_grad_norm": 0.4922628626227379, "chaser_new_fields": 72.3499984741211, "chaser_partner_vision": 0.5845000147819519, "chaser_return": 14.735000610351562, "chaser_subspace_norm": 50.69660186767578, "collisions_per_episode": 9.65000057220459, "entropy": 0.12092708237469196, "explorer_grad_norm": 0.7831171005964279, "explorer_new_fields": 80.7750015258789, "explorer_partner_vision": 0.5845000147819519, "explorer_return": 61.087501525878906, "explorer_subspace_norm": 41.08533477783203, "final_distance": 2.647909164428711, "max_param_abs": 1.584478497505188, "policy_loss": -0.006801864627753673, "value_loss": 1.211836338043213}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:44.264098+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005644.264113, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9536593473088901, "elapsed_seconds": 3433.1736503120046, "env_steps_per_second": 4077.859562602576, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.495822637313339, "eta_seconds": 16184.961494328021, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004649059142094458, "chaser_grad_norm": 0.6182895749807358, "chaser_new_fields": 69.70000457763672, "chaser_partner_vision": 0.6392500400543213, "chaser_return": 15.160000801086426, "chaser_subspace_norm": 58.159481048583984, "collisions_per_episode": 10.550000190734863, "entropy": 0.11344870924949646, "explorer_grad_norm": 2.2838667929172516, "explorer_new_fields": 77.80000305175781, "explorer_partner_vision": 0.6392500400543213, "explorer_return": 57.26250076293945, "explorer_subspace_norm": 45.38612365722656, "final_distance": 2.344717264175415, "max_param_abs": 1.6354731321334839, "policy_loss": -0.004198888045269911, "value_loss": 1.490211397409439}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:05.108374+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006145.1083872, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0889667859241794, "elapsed_seconds": 3920.2804293270456, "env_steps_per_second": 4081.3406817293826, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.355867143696718, "eta_seconds": 15681.121717308182, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0038515464057411464, "chaser_grad_norm": 2.0432513058185577, "chaser_new_fields": 69.7249984741211, "chaser_partner_vision": 0.6497500538825989, "chaser_return": 14.94250202178955, "chaser_subspace_norm": 67.1459732055664, "collisions_per_episode": 10.475000381469727, "entropy": 0.15305231511592865, "explorer_grad_norm": 0.8656620681285858, "explorer_new_fields": 79.0, "explorer_partner_vision": 0.6497500538825989, "explorer_return": 58.61249923706055, "explorer_subspace_norm": 46.93916702270508, "final_distance": 2.1592462062835693, "max_param_abs": 1.714181900024414, "policy_loss": -0.005022043010214361, "value_loss": 1.9083105325698853}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:12.215151+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006632.2151654, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2228155863666648, "elapsed_seconds": 4402.136110919993, "env_steps_per_second": 4088.9240010886933, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.2119203530407345, "eta_seconds": 15162.913270946645, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.004799993393154622, "chaser_grad_norm": 0.6318073645234108, "chaser_new_fields": 68.30000305175781, "chaser_partner_vision": 0.5782500505447388, "chaser_return": 11.950000762939453, "chaser_subspace_norm": 69.7387466430664, "collisions_per_episode": 7.800000190734863, "entropy": 0.13057992979884148, "explorer_grad_norm": 0.9818566143512726, "explorer_new_fields": 81.30000305175781, "explorer_partner_vision": 0.5782500505447388, "explorer_return": 64.3375015258789, "explorer_subspace_norm": 47.37717819213867, "final_distance": 2.072697401046753, "max_param_abs": 1.7917687892913818, "policy_loss": -0.0076305684354167624, "value_loss": 1.346177577972412}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:14.070833+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007114.0708442, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3568196338658325, "elapsed_seconds": 4884.550681916997, "env_steps_per_second": 4094.5424261931857, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.0704589015974975, "eta_seconds": 14653.65204575099, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0055760749853490665, "chaser_grad_norm": 0.5583522021770477, "chaser_new_fields": 63.79999923706055, "chaser_partner_vision": 0.5257500410079956, "chaser_return": 14.280001640319824, "chaser_subspace_norm": 67.18833923339844, "collisions_per_episode": 10.850000381469727, "entropy": 0.18042557314038277, "explorer_grad_norm": 1.708299607038498, "explorer_new_fields": 78.5250015258789, "explorer_partner_vision": 0.5257500410079956, "explorer_return": 58.875, "explorer_subspace_norm": 52.10688018798828, "final_distance": 2.4457335472106934, "max_param_abs": 1.8546721935272217, "policy_loss": -0.007979492247102371, "value_loss": 2.7333709597587585}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:16.485403+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007596.4854174, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4920222193025015, "elapsed_seconds": 5371.279989489005, "env_steps_per_second": 4095.857978554747, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9335131236156853, "eta_seconds": 14160.647245016467, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.003676087270697792, "chaser_grad_norm": 0.7119852304458618, "chaser_new_fields": 66.95000457763672, "chaser_partner_vision": 0.6000000238418579, "chaser_return": 11.480000495910645, "chaser_subspace_norm": 85.72869110107422, "collisions_per_episode": 7.550000190734863, "entropy": 0.14804701134562492, "explorer_grad_norm": 0.8053766936063766, "explorer_new_fields": 79.20000457763672, "explorer_partner_vision": 0.6000000238418579, "explorer_return": 61.01250076293945, "explorer_subspace_norm": 55.93586349487305, "final_distance": 2.1068623065948486, "max_param_abs": 1.9221315383911133, "policy_loss": -0.007040791582319628, "value_loss": 1.7720458507537842}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:23.214712+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008083.2147224, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6260300483552985, "elapsed_seconds": 5853.7081740790745, "env_steps_per_second": 4099.965233366926, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.79407011282903, "eta_seconds": 13658.652406184508, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.003633705058000203, "chaser_grad_norm": 0.6213411167263985, "chaser_new_fields": 70.80000305175781, "chaser_partner_vision": 0.640500009059906, "chaser_return": 13.239999771118164, "chaser_subspace_norm": 91.18304443359375, "collisions_per_episode": 8.550000190734863, "entropy": 0.14571617171168327, "explorer_grad_norm": 1.4613387286663055, "explorer_new_fields": 78.625, "explorer_partner_vision": 0.640500009059906, "explorer_return": 60.0625, "explorer_subspace_norm": 61.71382522583008, "final_distance": 2.0097997188568115, "max_param_abs": 1.9503538608551025, "policy_loss": -0.006107364140449434, "value_loss": 1.920810878276825}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:25.642896+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008565.6429105, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.763679194277243, "elapsed_seconds": 6349.245099398075, "env_steps_per_second": 4094.975007732001, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.663026018883505, "eta_seconds": 13186.893667980617, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.005337592199401442, "chaser_grad_norm": 0.6554550528526306, "chaser_new_fields": 69.0999984741211, "chaser_partner_vision": 0.6567500233650208, "chaser_return": 12.15999984741211, "chaser_subspace_norm": 98.39765167236328, "collisions_per_episode": 7.800000190734863, "entropy": 0.135175671428442, "explorer_grad_norm": 0.9496551603078842, "explorer_new_fields": 77.05000305175781, "explorer_partner_vision": 0.6567500233650208, "explorer_return": 58.712501525878906, "explorer_subspace_norm": 63.310585021972656, "final_distance": 1.7082542181015015, "max_param_abs": 1.975804090499878, "policy_loss": -0.0071361744822979745, "value_loss": 2.11739444732666}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:41.179823+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009061.179837, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8980078800480502, "elapsed_seconds": 6832.828368172981, "env_steps_per_second": 4097.863796846235, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5248717772320934, "eta_seconds": 12689.538398035536, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.006107125011723724, "chaser_grad_norm": 0.7649366557598114, "chaser_new_fields": 69.5250015258789, "chaser_partner_vision": 0.690250039100647, "chaser_return": 14.830000877380371, "chaser_subspace_norm": 106.89220428466797, "collisions_per_episode": 10.25, "entropy": 0.14211685955524445, "explorer_grad_norm": 0.9025948792695999, "explorer_new_fields": 76.45000457763672, "explorer_partner_vision": 0.690250039100647, "explorer_return": 56.400001525878906, "explorer_subspace_norm": 65.74281311035156, "final_distance": 2.0601754188537598, "max_param_abs": 1.9993696212768555, "policy_loss": -0.00802169600875402, "value_loss": 2.2889803647994995}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:44.763090+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009544.7631044, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.035567426063062, "elapsed_seconds": 7328.042733827024, "env_steps_per_second": 4093.862589189991, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3926123767717704, "eta_seconds": 12213.404556378373, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00326219870614608, "chaser_grad_norm": 0.7246690765023232, "chaser_new_fields": 70.375, "chaser_partner_vision": 0.6322500109672546, "chaser_return": 15.135001182556152, "chaser_subspace_norm": 114.6466064453125, "collisions_per_episode": 10.40000057220459, "entropy": 0.13015475124120712, "explorer_grad_norm": 0.9277182966470718, "explorer_new_fields": 76.4000015258789, "explorer_partner_vision": 0.6322500109672546, "explorer_return": 55.61249923706055, "explorer_subspace_norm": 64.73603057861328, "final_distance": 1.9015884399414062, "max_param_abs": 2.0542945861816406, "policy_loss": -0.006135252587512241, "value_loss": 2.171432137489319}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:33:59.977455+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010039.9774663, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1719310500927773, "elapsed_seconds": 7818.951780333999, "env_steps_per_second": 4092.620200125223, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.257896575139166, "eta_seconds": 11728.427670500998, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.006349712928354734, "chaser_grad_norm": 1.231883630156517, "chaser_new_fields": 69.07500457763672, "chaser_partner_vision": 0.6492500305175781, "chaser_return": 12.52500057220459, "chaser_subspace_norm": 128.79739379882812, "collisions_per_episode": 8.25, "entropy": 0.11546356603503227, "explorer_grad_norm": 0.7010753750801086, "explorer_new_fields": 77.05000305175781, "explorer_partner_vision": 0.6492500305175781, "explorer_return": 58.26250076293945, "explorer_subspace_norm": 72.56559753417969, "final_distance": 2.002412796020508, "max_param_abs": 2.1060523986816406, "policy_loss": -0.004033066152597797, "value_loss": 2.046307146549225}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:10.886505+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010530.8865192, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3084164061972277, "elapsed_seconds": 8310.29906231002, "env_steps_per_second": 4091.3088379937312, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1231516083844846, "eta_seconds": 11243.345790184145, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.005581053996799312, "chaser_grad_norm": 0.4815082401037216, "chaser_new_fields": 70.8499984741211, "chaser_partner_vision": 0.7225000262260437, "chaser_return": 13.035000801086426, "chaser_subspace_norm": 110.86450958251953, "collisions_per_episode": 8.40000057220459, "entropy": 0.12675759196281433, "explorer_grad_norm": 0.6857897192239761, "explorer_new_fields": 75.2249984741211, "explorer_partner_vision": 0.7225000262260437, "explorer_return": 55.337501525878906, "explorer_subspace_norm": 71.9049301147461, "final_distance": 2.1238675117492676, "max_param_abs": 2.145883798599243, "policy_loss": -0.008801496719943458, "value_loss": 2.2562089562416077}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:22.233786+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011022.233798, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4428988123591666, "elapsed_seconds": 8794.435724493, "env_steps_per_second": 4093.4974258482516, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.985765215105648, "eta_seconds": 10748.754774380333, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.007456515981825618, "chaser_grad_norm": 0.9439582675695419, "chaser_new_fields": 70.7249984741211, "chaser_partner_vision": 0.6387500166893005, "chaser_return": 13.635001182556152, "chaser_subspace_norm": 117.0467529296875, "collisions_per_episode": 9.0, "entropy": 0.11708638444542885, "explorer_grad_norm": 1.0359466671943665, "explorer_new_fields": 76.3499984741211, "explorer_partner_vision": 0.6387500166893005, "explorer_return": 56.837501525878906, "explorer_subspace_norm": 68.80552673339844, "final_distance": 1.9229882955551147, "max_param_abs": 2.177574396133423, "policy_loss": -0.0062263856809749996, "value_loss": 1.7954291105270386}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:26.370447+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011506.3704615, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5767561115569615, "elapsed_seconds": 9276.322001605062, "env_steps_per_second": 4096.451157411843, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8479935969840104, "eta_seconds": 10252.776949142437, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00759867347940002, "chaser_grad_norm": 0.644098736345768, "chaser_new_fields": 69.80000305175781, "chaser_partner_vision": 0.6632500290870667, "chaser_return": 12.457500457763672, "chaser_subspace_norm": 151.96078491210938, "collisions_per_episode": 7.925000190734863, "entropy": 0.11262200959026814, "explorer_grad_norm": 1.0923659428954124, "explorer_new_fields": 78.0250015258789, "explorer_partner_vision": 0.6632500290870667, "explorer_return": 58.79999923706055, "explorer_subspace_norm": 76.7490234375, "final_distance": 1.7065365314483643, "max_param_abs": 2.2497031688690186, "policy_loss": -0.006229564202961058, "value_loss": 1.5732687413692474}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:28.256723+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011988.2567356, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.710964675381385, "elapsed_seconds": 9759.472831372987, "env_steps_per_second": 4098.582033182698, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.710964675381385, "eta_seconds": 9759.472831372987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.005510369429606143, "chaser_grad_norm": 0.8583532124757767, "chaser_new_fields": 69.82500457763672, "chaser_partner_vision": 0.7382500171661377, "chaser_return": 12.07750129699707, "chaser_subspace_norm": 156.27415466308594, "collisions_per_episode": 7.675000190734863, "entropy": 0.12425081804394722, "explorer_grad_norm": 1.6032065749168396, "explorer_new_fields": 79.5250015258789, "explorer_partner_vision": 0.7382500171661377, "explorer_return": 61.10000228881836, "explorer_subspace_norm": 71.80419158935547, "final_distance": 1.832790732383728, "max_param_abs": 2.271259307861328, "policy_loss": -0.00596293076313259, "value_loss": 1.8261153995990753}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:31.407553+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012471.4075685, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.846690764761394, "elapsed_seconds": 10248.086753141019, "env_steps_per_second": 4098.325961880356, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5755773585936423, "eta_seconds": 9272.078490937112, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0022896163755325993, "chaser_grad_norm": 0.34782107174396515, "chaser_new_fields": 73.7249984741211, "chaser_partner_vision": 0.7317500114440918, "chaser_return": 12.330000877380371, "chaser_subspace_norm": 152.9884796142578, "collisions_per_episode": 7.050000190734863, "entropy": 0.0966180395334959, "explorer_grad_norm": 0.5978739336133003, "explorer_new_fields": 78.20000457763672, "explorer_partner_vision": 0.7317500114440918, "explorer_return": 58.48749923706055, "explorer_subspace_norm": 74.78372192382812, "final_distance": 1.896437644958496, "max_param_abs": 2.269610643386841, "policy_loss": -0.006409946721042559, "value_loss": 1.438534528017044}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:22:40.021474+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012960.0214865, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.981165438270287, "elapsed_seconds": 10732.195577773033, "env_steps_per_second": 4099.8134707055115, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.43913535858478, "eta_seconds": 8780.887290905208, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004979065920505121, "chaser_grad_norm": 0.6363644301891327, "chaser_new_fields": 72.4000015258789, "chaser_partner_vision": 0.7770000100135803, "chaser_return": 12.607500076293945, "chaser_subspace_norm": 167.10122680664062, "collisions_per_episode": 7.675000190734863, "entropy": 0.10450165718793869, "explorer_grad_norm": 0.6356114894151688, "explorer_new_fields": 78.30000305175781, "explorer_partner_vision": 0.7770000100135803, "explorer_return": 58.13750076293945, "explorer_subspace_norm": 75.46092224121094, "final_distance": 1.8890907764434814, "max_param_abs": 2.3103115558624268, "policy_loss": -0.007887110084290327, "value_loss": 1.5648295283317566}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:30:44.130300+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013444.1303153, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1175751288555653, "elapsed_seconds": 11223.270463880035, "env_steps_per_second": 4098.627057776275, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.304294660458461, "eta_seconds": 8295.46077765046, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.006847067117048411, "chaser_grad_norm": 0.6606469005346298, "chaser_new_fields": 54.75, "chaser_partner_vision": 0.9362500309944153, "chaser_return": 12.930000305175781, "chaser_subspace_norm": 161.50027465820312, "collisions_per_episode": 11.350000381469727, "entropy": 0.19372183457016945, "explorer_grad_norm": 0.8121878504753113, "explorer_new_fields": 66.9000015258789, "explorer_partner_vision": 0.9362500309944153, "explorer_return": 40.32500076293945, "explorer_subspace_norm": 84.08326721191406, "final_distance": 2.3542230129241943, "max_param_abs": 2.3364620208740234, "policy_loss": -0.008452974626806053, "value_loss": 3.405247151851654}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:38:55.205187+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013935.2052002, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.254728750804465, "elapsed_seconds": 11717.023502896074, "env_steps_per_second": 4096.603543394441, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1698191672029767, "eta_seconds": 7811.349001930716, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0028354159808974266, "chaser_grad_norm": 0.5318872407078743, "chaser_new_fields": 58.400001525878906, "chaser_partner_vision": 0.9550000429153442, "chaser_return": 13.017500877380371, "chaser_subspace_norm": 139.70175170898438, "collisions_per_episode": 10.77500057220459, "entropy": 0.18797292560338974, "explorer_grad_norm": 1.1622335463762283, "explorer_new_fields": 69.125, "explorer_partner_vision": 0.9550000429153442, "explorer_return": 43.76250076293945, "explorer_subspace_norm": 76.24115753173828, "final_distance": 1.7495298385620117, "max_param_abs": 2.3502728939056396, "policy_loss": -0.007433881672113785, "value_loss": 3.344414234161377}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:08.958226+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014428.9582405, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.390079421494447, "elapsed_seconds": 12204.28591738001, "env_steps_per_second": 4096.9213879851395, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0340476528966684, "eta_seconds": 7322.571550428006, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00461313084024173, "chaser_grad_norm": 0.852738730609417, "chaser_new_fields": 54.0, "chaser_partner_vision": 0.9632500410079956, "chaser_return": 14.415000915527344, "chaser_subspace_norm": 158.68968200683594, "collisions_per_episode": 12.90000057220459, "entropy": 0.17376044020056725, "explorer_grad_norm": 0.8431649655103683, "explorer_new_fields": 66.375, "explorer_partner_vision": 0.9632500410079956, "explorer_return": 38.38750076293945, "explorer_subspace_norm": 82.70452117919922, "final_distance": 1.754865288734436, "max_param_abs": 2.399944305419922, "policy_loss": -0.005140919778497732, "value_loss": 3.38133704662323}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:16.220639+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014916.2206507, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.525692128422217, "elapsed_seconds": 12692.491662319982, "env_steps_per_second": 4096.910313864665, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.898449607611963, "eta_seconds": 6834.418587403067, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.005451803702675839, "chaser_grad_norm": 0.7529477775096893, "chaser_new_fields": 54.400001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 15.90000057220459, "chaser_subspace_norm": 162.20318603515625, "collisions_per_episode": 14.300000190734863, "entropy": 0.14639435335993767, "explorer_grad_norm": 0.7564421147108078, "explorer_new_fields": 67.125, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 37.650001525878906, "explorer_subspace_norm": 91.4089126586914, "final_distance": 1.6693363189697266, "max_param_abs": 2.4423253536224365, "policy_loss": -0.005958304470368603, "value_loss": 2.7669906616210938}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:24.426385+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015404.4263988, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6598307460680695, "elapsed_seconds": 13175.39068584505, "env_steps_per_second": 4098.5501900915, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7621407295883296, "eta_seconds": 6343.7066265179865, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.005560478827177917, "chaser_grad_norm": 0.42106715589761734, "chaser_new_fields": 52.60000228881836, "chaser_partner_vision": 0.9842500686645508, "chaser_return": 18.457500457763672, "chaser_subspace_norm": 169.46694946289062, "collisions_per_episode": 16.825000762939453, "entropy": 0.15263331308960915, "explorer_grad_norm": 0.6271104067564011, "explorer_new_fields": 62.29999923706055, "explorer_partner_vision": 0.9842500686645508, "explorer_return": 29.524999618530273, "explorer_subspace_norm": 88.21095275878906, "final_distance": 1.814388632774353, "max_param_abs": 2.4204506874084473, "policy_loss": -0.006235027218252753, "value_loss": 3.5767462849617004}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:27.325408+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015887.325419, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7966319625327984, "elapsed_seconds": 13667.875065118074, "env_steps_per_second": 4097.198703763263, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6271279839426276, "eta_seconds": 5857.66074219346, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0043260492873571366, "chaser_grad_norm": 0.5565928593277931, "chaser_new_fields": 56.82500076293945, "chaser_partner_vision": 0.9800000190734863, "chaser_return": 16.27250099182129, "chaser_subspace_norm": 154.13037109375, "collisions_per_episode": 14.27500057220459, "entropy": 0.15590067952871323, "explorer_grad_norm": 1.1441697925329208, "explorer_new_fields": 68.625, "explorer_partner_vision": 0.9800000190734863, "explorer_return": 39.95000076293945, "explorer_subspace_norm": 87.79729461669922, "final_distance": 1.7146209478378296, "max_param_abs": 2.4479570388793945, "policy_loss": -0.005616793930528274, "value_loss": 2.9997896552085876}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:19:39.809787+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016379.809801, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.934848515956126, "elapsed_seconds": 14165.454657442053, "env_steps_per_second": 4094.467943500052, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.492528747431634, "eta_seconds": 5373.103490753882, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0035942451398636166, "chaser_grad_norm": 1.0028918832540512, "chaser_new_fields": 52.47500228881836, "chaser_partner_vision": 0.9752500653266907, "chaser_return": 19.66750144958496, "chaser_subspace_norm": 168.74253845214844, "collisions_per_episode": 18.225000381469727, "entropy": 0.16648008674383163, "explorer_grad_norm": 1.0500975996255875, "explorer_new_fields": 58.900001525878906, "explorer_partner_vision": 0.9752500653266907, "explorer_return": 24.475000381469727, "explorer_subspace_norm": 84.93462371826172, "final_distance": 1.755316972732544, "max_param_abs": 2.4847402572631836, "policy_loss": -0.004450086853466928, "value_loss": 3.64981472492218}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:27:57.389379+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016877.389391, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.073471720393605, "elapsed_seconds": 14664.498193416977, "env_steps_per_second": 4091.5140230938505, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3578239067978684, "eta_seconds": 4888.166064472326, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0036971174051032563, "chaser_grad_norm": 1.4342991411685944, "chaser_new_fields": 54.900001525878906, "chaser_partner_vision": 0.968250036239624, "chaser_return": 16.48750114440918, "chaser_subspace_norm": 182.44105529785156, "collisions_per_episode": 14.625, "entropy": 0.20962896198034286, "explorer_grad_norm": 1.0151912868022919, "explorer_new_fields": 63.900001525878906, "explorer_partner_vision": 0.968250036239624, "explorer_return": 34.5625, "explorer_subspace_norm": 71.98743438720703, "final_distance": 1.9809166193008423, "max_param_abs": 2.5495941638946533, "policy_loss": -0.004625928693918979, "value_loss": 3.501450538635254}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:16.432916+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017376.4329312, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.21044605074583, "elapsed_seconds": 15157.605782684986, "env_steps_per_second": 4090.35575201623, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.222387563119757, "eta_seconds": 4400.595227231125, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00615208081003793, "chaser_grad_norm": 0.4877852499485016, "chaser_new_fields": 57.85000228881836, "chaser_partner_vision": 0.9727500677108765, "chaser_return": 13.045000076293945, "chaser_subspace_norm": 184.0464630126953, "collisions_per_episode": 11.050000190734863, "entropy": 0.18670667335391045, "explorer_grad_norm": 0.7557024359703064, "explorer_new_fields": 69.05000305175781, "explorer_partner_vision": 0.9727500677108765, "explorer_return": 44.78750228881836, "explorer_subspace_norm": 70.16930389404297, "final_distance": 1.9953289031982422, "max_param_abs": 2.586437225341797, "policy_loss": -0.006206624491142065, "value_loss": 2.8411368131637573}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:29.540504+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017869.5405142, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.345058956344179, "elapsed_seconds": 15642.212242839043, "env_steps_per_second": 4091.4928787837543, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0862647390860447, "eta_seconds": 3910.5530607097608, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.006657696801567248, "chaser_grad_norm": 0.4136536568403244, "chaser_new_fields": 57.25, "chaser_partner_vision": 0.9660000205039978, "chaser_return": 14.58750057220459, "chaser_subspace_norm": 188.2316436767578, "collisions_per_episode": 12.52500057220459, "entropy": 0.1750296913087368, "explorer_grad_norm": 0.6128263473510742, "explorer_new_fields": 65.5999984741211, "explorer_partner_vision": 0.9660000205039978, "explorer_return": 36.8125, "explorer_subspace_norm": 73.95885467529297, "final_distance": 1.8617504835128784, "max_param_abs": 2.6188292503356934, "policy_loss": -0.006537529943514642, "value_loss": 3.1692182421684265}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:34.146967+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018354.146981, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.483289353779724, "elapsed_seconds": 16139.841673607007, "env_steps_per_second": 4089.2594447148635, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9510007720138808, "eta_seconds": 3423.602779249971, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0018832554883090324, "chaser_grad_norm": 0.549080565571785, "chaser_new_fields": 55.47500228881836, "chaser_partner_vision": 0.9865000247955322, "chaser_return": 14.6875, "chaser_subspace_norm": 182.03871154785156, "collisions_per_episode": 12.925000190734863, "entropy": 0.1828586608171463, "explorer_grad_norm": 1.497645691037178, "explorer_new_fields": 64.45000457763672, "explorer_partner_vision": 0.9865000247955322, "explorer_return": 37.025001525878906, "explorer_subspace_norm": 68.02664184570312, "final_distance": 1.7270240783691406, "max_param_abs": 2.6212704181671143, "policy_loss": -0.0035793553958569646, "value_loss": 3.2441684007644653}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:51.776397+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018851.7764125, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.621179112262778, "elapsed_seconds": 16636.244804146, "env_steps_per_second": 4087.4608903959734, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8155021962816668, "eta_seconds": 2935.8079066140003, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.005208360076103258, "chaser_grad_norm": 0.6875676587224007, "chaser_new_fields": 49.92499923706055, "chaser_partner_vision": 0.9800000190734863, "chaser_return": 17.970001220703125, "chaser_subspace_norm": 190.2951202392578, "collisions_per_episode": 17.200000762939453, "entropy": 0.18518804758787155, "explorer_grad_norm": 0.8631334602832794, "explorer_new_fields": 59.47500228881836, "explorer_partner_vision": 0.9800000190734863, "explorer_return": 25.212499618530273, "explorer_subspace_norm": 83.83126831054688, "final_distance": 1.5211904048919678, "max_param_abs": 2.662728786468506, "policy_loss": -0.0055554613098536265, "value_loss": 3.823395609855652}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:09:08.179528+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019348.1795428, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.756664194638336, "elapsed_seconds": 17123.99110069801, "env_steps_per_second": 4087.8320707108196, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6795234563769051, "eta_seconds": 2446.2844429568586, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.009193059900493239, "chaser_grad_norm": 0.5828518643975258, "chaser_new_fields": 52.97500228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 15.027499198913574, "chaser_subspace_norm": 189.95901489257812, "collisions_per_episode": 13.824999809265137, "entropy": 0.21082253754138947, "explorer_grad_norm": 0.8342098742723465, "explorer_new_fields": 61.85000228881836, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 31.25, "explorer_subspace_norm": 80.66325378417969, "final_distance": 1.5274208784103394, "max_param_abs": 2.682708263397217, "policy_loss": -0.005754835532030711, "value_loss": 3.749134600162506}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:17:15.925822+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019835.9258335, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.895630027611688, "elapsed_seconds": 17624.268099402077, "env_steps_per_second": 4085.276029274809, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5439588919568542, "eta_seconds": 1958.252011044675, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0020387819210156977, "chaser_grad_norm": 0.5956076234579086, "chaser_new_fields": 57.85000228881836, "chaser_partner_vision": 0.9807500243186951, "chaser_return": 12.002501487731934, "chaser_subspace_norm": 197.14990234375, "collisions_per_episode": 9.975000381469727, "entropy": 0.1588515229523182, "explorer_grad_norm": 0.4954075366258621, "explorer_new_fields": 67.0250015258789, "explorer_partner_vision": 0.9807500243186951, "explorer_return": 41.61249923706055, "explorer_subspace_norm": 75.98540496826172, "final_distance": 1.7454546689987183, "max_param_abs": 2.6986279487609863, "policy_loss": -0.004600939356748768, "value_loss": 2.678371548652649}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:36.202821+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020336.202835, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.031490470426118, "elapsed_seconds": 18113.365693534026, "env_steps_per_second": 4085.380997216656, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40795868679130687, "eta_seconds": 1468.6512724487047, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.002321154035240802, "chaser_grad_norm": 0.6193432807922363, "chaser_new_fields": 58.70000076293945, "chaser_partner_vision": 0.98375004529953, "chaser_return": 12.072500228881836, "chaser_subspace_norm": 210.81329345703125, "collisions_per_episode": 9.975000381469727, "entropy": 0.1777302622795105, "explorer_grad_norm": 0.8054322004318237, "explorer_new_fields": 69.05000305175781, "explorer_partner_vision": 0.98375004529953, "explorer_return": 44.20000076293945, "explorer_subspace_norm": 77.97282409667969, "final_distance": 2.1908769607543945, "max_param_abs": 2.6941137313842773, "policy_loss": -0.0045320950057359255, "value_loss": 2.851964235305786}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:45.300416+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020825.3004298, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.169450981867493, "elapsed_seconds": 18610.023534722975, "env_steps_per_second": 4083.820735540586, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27207636746671016, "eta_seconds": 979.4749228801566, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0038528375844855134, "chaser_grad_norm": 0.5149100199341774, "chaser_new_fields": 62.35000228881836, "chaser_partner_vision": 0.986750066280365, "chaser_return": 13.827499389648438, "chaser_subspace_norm": 242.19839477539062, "collisions_per_episode": 10.875, "entropy": 0.13560564443469048, "explorer_grad_norm": 0.7652166336774826, "explorer_new_fields": 71.82500457763672, "explorer_partner_vision": 0.986750066280365, "explorer_return": 47.650001525878906, "explorer_subspace_norm": 84.7203140258789, "final_distance": 1.8313133716583252, "max_param_abs": 2.7527127265930176, "policy_loss": -0.00582649988121986, "value_loss": 2.2465010285377502}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:42:01.958257+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021321.9582722, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.303272566010289, "elapsed_seconds": 19091.781237637042, "env_steps_per_second": 4085.527642975126, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13598134784641766, "eta_seconds": 489.5328522471036, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.005702570685564012, "chaser_grad_norm": 0.8132647722959518, "chaser_new_fields": 61.67499923706055, "chaser_partner_vision": 0.9515000581741333, "chaser_return": 8.920000076293945, "chaser_subspace_norm": 221.4047393798828, "collisions_per_episode": 6.349999904632568, "entropy": 0.17935073748230934, "explorer_grad_norm": 0.7646839916706085, "explorer_new_fields": 71.32500457763672, "explorer_partner_vision": 0.9515000581741333, "explorer_return": 51.97500228881836, "explorer_subspace_norm": 90.32686614990234, "final_distance": 2.1570069789886475, "max_param_abs": 2.777965545654297, "policy_loss": -0.005222004773088429, "value_loss": 2.7299355268478394}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:50:03.715959+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021803.71597, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.435660540801958, "elapsed_seconds": 19568.37794688705, "env_steps_per_second": 4088.2284784736826, "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-transformer/social/seed_0002/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:00.312668+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022280.3126786, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.435672476867783, "elapsed_seconds": 19568.42091672402, "env_steps_per_second": 4088.2195012285606, "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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0025767051566676843, "chaser_grad_norm": 0.482593297958374, "chaser_new_fields": 56.32500076293945, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 15.255000114440918, "chaser_subspace_norm": 223.2849578857422, "collisions_per_episode": 13.199999809265137, "entropy": 0.16332043334841728, "explorer_grad_norm": 0.6213180869817734, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 33.775001525878906, "explorer_subspace_norm": 86.63322448730469, "final_distance": 1.4801658391952515, "max_param_abs": 2.83324933052063, "policy_loss": -0.004363851029381749, "value_loss": 3.7607646584510803}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:58:00.355640+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022280.355656, "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..5f0fdc28dc31bb2d484913ed6dba2b81501ccbcb --- /dev/null +++ b/social/seed_0002/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.435672476867783, + "elapsed_seconds": 19568.42091672402, + "env_steps_per_second": 4088.2195012285606, + "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-transformer/social/seed_0002/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0025767051566676843, + "chaser_grad_norm": 0.482593297958374, + "chaser_new_fields": 56.32500076293945, + "chaser_partner_vision": 0.9642500281333923, + "chaser_return": 15.255000114440918, + "chaser_subspace_norm": 223.2849578857422, + "collisions_per_episode": 13.199999809265137, + "entropy": 0.16332043334841728, + "explorer_grad_norm": 0.6213180869817734, + "explorer_new_fields": 62.5, + "explorer_partner_vision": 0.9642500281333923, + "explorer_return": 33.775001525878906, + "explorer_subspace_norm": 86.63322448730469, + "final_distance": 1.4801658391952515, + "max_param_abs": 2.83324933052063, + "policy_loss": -0.004363851029381749, + "value_loss": 3.7607646584510803 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:58:00.355640+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022280.355656, + "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..65e5c403c4f922c442d7a0ca5f2c30188fae3211 --- /dev/null +++ b/social/seed_0003/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:42e883b23fb099df0bbca6fff98d93505336beb9d24fcca6844a6a86809443c2 +size 29760368 diff --git a/social/seed_0003/attempt_01/config.json b/social/seed_0003/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..5b6848c728c52247623d95f8afc09aa87dab22bb --- /dev/null +++ b/social/seed_0003/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0003/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..825e6e696bec66b080bc3aa0b9d65eba6f0ef592 --- /dev/null +++ b/social/seed_0003/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.000830947157196028, "elapsed_seconds": 2.9914097659057006, "env_steps_per_second": 1337.1621787123943, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 16.618112196763363, "eta_seconds": 59825.20390834811, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005422881372360688, "chaser_grad_norm": 2.819717012345791, "chaser_new_fields": 32.17499923706055, "chaser_partner_vision": 0.34175002574920654, "chaser_return": -1.8450002670288086, "chaser_subspace_norm": 0.031720127910375595, "collisions_per_episode": 1.600000023841858, "entropy": 2.755132257938385, "explorer_grad_norm": 9.636096179485321, "explorer_new_fields": 31.225000381469727, "explorer_partner_vision": 0.34175002574920654, "explorer_return": -4.224999904632568, "explorer_subspace_norm": 0.038686346262693405, "final_distance": 5.089163303375244, "max_param_abs": 1.0012006759643555, "policy_loss": -0.010342364998564335, "value_loss": 7.5233811140060425}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.875178+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.8751907, "update": 1} +{"device": "cuda", "elapsed_hours": 0.1356185472391533, "elapsed_seconds": 488.2267700609518, "env_steps_per_second": 4096.457061849176, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.289123342326978, "eta_seconds": 19040.84403237712, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0023989090196749885, "chaser_grad_norm": 1.1969573646783829, "chaser_new_fields": 59.400001525878906, "chaser_partner_vision": 0.5020000338554382, "chaser_return": 12.637499809265137, "chaser_subspace_norm": 7.698081016540527, "collisions_per_episode": 9.975000381469727, "entropy": 0.6663153171539307, "explorer_grad_norm": 1.9852930903434753, "explorer_new_fields": 65.57500457763672, "explorer_partner_vision": 0.5020000338554382, "explorer_return": 41.01250076293945, "explorer_subspace_norm": 8.45530891418457, "final_distance": 4.368406295776367, "max_param_abs": 1.2258557081222534, "policy_loss": -0.00611149521078147, "value_loss": 3.6279552578926086}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:40:00.110539+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003200.11055, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2716397749086092, "elapsed_seconds": 977.903189670993, "env_steps_per_second": 4090.384449350007, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.161155723263574, "eta_seconds": 18580.160603748867, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.007366140425032142, "chaser_grad_norm": 0.6448420062661171, "chaser_new_fields": 52.97500228881836, "chaser_partner_vision": 0.640500009059906, "chaser_return": 15.445001602172852, "chaser_subspace_norm": 14.25063419342041, "collisions_per_episode": 13.850000381469727, "entropy": 0.5230071842670441, "explorer_grad_norm": 1.8020373284816742, "explorer_new_fields": 63.525001525878906, "explorer_partner_vision": 0.640500009059906, "explorer_return": 34.95000076293945, "explorer_subspace_norm": 13.386552810668945, "final_distance": 3.353379964828491, "max_param_abs": 1.3246737718582153, "policy_loss": -0.01077506886053925, "value_loss": 3.96921169757843}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:09.786958+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003689.7869725, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.40855010376360346, "elapsed_seconds": 1470.7803735489724, "env_steps_per_second": 4079.467001264155, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.0387846130844425, "eta_seconds": 18139.624607103993, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0039427688984874365, "chaser_grad_norm": 1.7115550339221954, "chaser_new_fields": 54.47500228881836, "chaser_partner_vision": 0.625, "chaser_return": 9.482501029968262, "chaser_subspace_norm": 25.27991485595703, "collisions_per_episode": 8.02500057220459, "entropy": 0.3656124100089073, "explorer_grad_norm": 2.1406101882457733, "explorer_new_fields": 69.0, "explorer_partner_vision": 0.625, "explorer_return": 46.75, "explorer_subspace_norm": 18.77017593383789, "final_distance": 2.2851192951202393, "max_param_abs": 1.4050453901290894, "policy_loss": -0.006032475594430764, "value_loss": 2.854288935661316}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:22.664141+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004182.6641526, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5435703857436116, "elapsed_seconds": 1956.853388677002, "env_steps_per_second": 4088.1958997493803, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.892133471692505, "eta_seconds": 17611.680498093017, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0028110619108971946, "chaser_grad_norm": 1.4864625036716461, "chaser_new_fields": 52.72500228881836, "chaser_partner_vision": 0.6642500162124634, "chaser_return": 10.289999961853027, "chaser_subspace_norm": 29.055410385131836, "collisions_per_episode": 9.199999809265137, "entropy": 0.3684553802013397, "explorer_grad_norm": 1.3596716225147247, "explorer_new_fields": 66.55000305175781, "explorer_partner_vision": 0.6642500162124634, "explorer_return": 42.375, "explorer_subspace_norm": 21.060224533081055, "final_distance": 2.5567433834075928, "max_param_abs": 1.4603949785232544, "policy_loss": -0.007349306887022067, "value_loss": 3.276752471923828}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:28.737158+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004668.7371721, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6800497004516526, "elapsed_seconds": 2448.1789216259494, "env_steps_per_second": 4084.6687763157993, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.760347903161568, "eta_seconds": 17137.252451381646, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0027782528449904476, "chaser_grad_norm": 1.3182986378669739, "chaser_new_fields": 52.85000228881836, "chaser_partner_vision": 0.7160000205039978, "chaser_return": 14.190001487731934, "chaser_subspace_norm": 34.536293029785156, "collisions_per_episode": 12.75, "entropy": 0.3412190452218056, "explorer_grad_norm": 1.5054287910461426, "explorer_new_fields": 65.1500015258789, "explorer_partner_vision": 0.7160000205039978, "explorer_return": 38.01250076293945, "explorer_subspace_norm": 27.5476016998291, "final_distance": 2.3847432136535645, "max_param_abs": 1.5125839710235596, "policy_loss": -0.007277713701135724, "value_loss": 3.688839852809906}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:40.062688+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005160.0626998, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8164258475769182, "elapsed_seconds": 2939.1330512769055, "env_steps_per_second": 4082.836602033584, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.626413136269203, "eta_seconds": 16655.087290569132, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.007106295309197819, "chaser_grad_norm": 1.1358559876680374, "chaser_new_fields": 55.70000076293945, "chaser_partner_vision": 0.7897500395774841, "chaser_return": 14.167500495910645, "chaser_subspace_norm": 42.01456832885742, "collisions_per_episode": 12.27500057220459, "entropy": 0.30694832652807236, "explorer_grad_norm": 1.507403388619423, "explorer_new_fields": 66.42500305175781, "explorer_partner_vision": 0.7897500395774841, "explorer_return": 39.29999923706055, "explorer_subspace_norm": 26.433454513549805, "final_distance": 2.3832523822784424, "max_param_abs": 1.5849307775497437, "policy_loss": -0.0061051797866724655, "value_loss": 3.4651512503623962}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:20:51.016820+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005651.016836, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9498820066291632, "elapsed_seconds": 3419.5752238649875, "env_steps_per_second": 4094.0757501969642, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.478015174108912, "eta_seconds": 16120.854626792083, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004821792918463075, "chaser_grad_norm": 1.2193410843610764, "chaser_new_fields": 51.25, "chaser_partner_vision": 0.8452500104904175, "chaser_return": 13.664999961853027, "chaser_subspace_norm": 42.73237609863281, "collisions_per_episode": 12.600000381469727, "entropy": 0.3655410185456276, "explorer_grad_norm": 1.0601339489221573, "explorer_new_fields": 59.29999923706055, "explorer_partner_vision": 0.8452500104904175, "explorer_return": 29.725000381469727, "explorer_subspace_norm": 28.79182243347168, "final_distance": 2.062945604324341, "max_param_abs": 1.6215722560882568, "policy_loss": -0.007752364492041863, "value_loss": 3.7555503249168396}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:28:51.458991+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006131.4590032, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0854837528146972, "elapsed_seconds": 3907.7415101329098, "env_steps_per_second": 4094.436635205129, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.341935011258789, "eta_seconds": 15630.966040531639, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0037071942276113434, "chaser_grad_norm": 0.646816298365593, "chaser_new_fields": 51.22500228881836, "chaser_partner_vision": 0.9417500495910645, "chaser_return": 18.04250144958496, "chaser_subspace_norm": 59.22454071044922, "collisions_per_episode": 16.975000381469727, "entropy": 0.3050040155649185, "explorer_grad_norm": 1.0369515866041183, "explorer_new_fields": 59.875, "explorer_partner_vision": 0.9417500495910645, "explorer_return": 25.475000381469727, "explorer_subspace_norm": 32.796836853027344, "final_distance": 2.049037456512451, "max_param_abs": 1.6850343942642212, "policy_loss": -0.00947592548936882, "value_loss": 3.826480269432068}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:36:59.625278+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006619.625293, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2191129748824945, "elapsed_seconds": 4388.80670957698, "env_steps_per_second": 4101.3426179652715, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.199166913484148, "eta_seconds": 15117.00088854293, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.004254628111559633, "chaser_grad_norm": 0.713115282356739, "chaser_new_fields": 47.54999923706055, "chaser_partner_vision": 0.9462500214576721, "chaser_return": 18.797500610351562, "chaser_subspace_norm": 66.86422729492188, "collisions_per_episode": 18.075000762939453, "entropy": 0.25790710747241974, "explorer_grad_norm": 1.3013441413640976, "explorer_new_fields": 54.79999923706055, "explorer_partner_vision": 0.9462500214576721, "explorer_return": 19.11250114440918, "explorer_subspace_norm": 34.926326751708984, "final_distance": 1.6684455871582031, "max_param_abs": 1.729909062385559, "policy_loss": -0.006359167935315213, "value_loss": 4.597017884254456}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:00.690477+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007100.6904898, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3557651975846965, "elapsed_seconds": 4880.754711304908, "env_steps_per_second": 4097.726926058295, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.067295592754089, "eta_seconds": 14642.264133914721, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0035724663226943676, "chaser_grad_norm": 0.9525994807481766, "chaser_new_fields": 51.650001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 17.957502365112305, "chaser_subspace_norm": 75.90569305419922, "collisions_per_episode": 16.725000381469727, "entropy": 0.24491872265934944, "explorer_grad_norm": 0.77274589240551, "explorer_new_fields": 59.900001525878906, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 26.875, "explorer_subspace_norm": 35.855552673339844, "final_distance": 1.8730367422103882, "max_param_abs": 1.8098558187484741, "policy_loss": -0.00699548556190166, "value_loss": 3.7204041481018066}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:12.638481+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007592.638496, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4917616512360918, "elapsed_seconds": 5370.34194444993, "env_steps_per_second": 4096.573407720577, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.932826171440605, "eta_seconds": 14158.174217186179, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0061281431817512555, "chaser_grad_norm": 1.0439439564943314, "chaser_new_fields": 53.17499923706055, "chaser_partner_vision": 0.9687500596046448, "chaser_return": 19.610000610351562, "chaser_subspace_norm": 78.6053695678711, "collisions_per_episode": 18.05000114440918, "entropy": 0.24620985612273216, "explorer_grad_norm": 0.8680069297552109, "explorer_new_fields": 60.025001525878906, "explorer_partner_vision": 0.9687500596046448, "explorer_return": 24.899999618530273, "explorer_subspace_norm": 39.32476806640625, "final_distance": 1.716272234916687, "max_param_abs": 1.859700083732605, "policy_loss": -0.007670742777122541, "value_loss": 3.845792055130005}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:22.225712+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008082.2257247, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.628630118923029, "elapsed_seconds": 5863.068428122904, "env_steps_per_second": 4093.4197330669294, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.800136944153734, "eta_seconds": 13680.492998953443, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.005777062335269534, "chaser_grad_norm": 1.063619390130043, "chaser_new_fields": 50.400001525878906, "chaser_partner_vision": 0.9657500386238098, "chaser_return": 19.850000381469727, "chaser_subspace_norm": 84.4339828491211, "collisions_per_episode": 18.700000762939453, "entropy": 0.251777783036232, "explorer_grad_norm": 1.0014899224042892, "explorer_new_fields": 56.35000228881836, "explorer_partner_vision": 0.9657500386238098, "explorer_return": 19.4375, "explorer_subspace_norm": 42.677581787109375, "final_distance": 1.5605347156524658, "max_param_abs": 1.9248143434524536, "policy_loss": -0.006608555803069294, "value_loss": 4.019231915473938}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:09:34.952197+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008574.9522114, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7651538922749912, "elapsed_seconds": 6354.5540121899685, "env_steps_per_second": 4091.5538604478124, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6660888531865203, "eta_seconds": 13197.919871471473, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0020765578990311617, "chaser_grad_norm": 0.4205848053097725, "chaser_new_fields": 55.150001525878906, "chaser_partner_vision": 0.971000075340271, "chaser_return": 16.325000762939453, "chaser_subspace_norm": 98.29732513427734, "collisions_per_episode": 14.699999809265137, "entropy": 0.20979038253426552, "explorer_grad_norm": 0.7758474797010422, "explorer_new_fields": 62.72500228881836, "explorer_partner_vision": 0.971000075340271, "explorer_return": 31.600000381469727, "explorer_subspace_norm": 41.91112518310547, "final_distance": 1.492792010307312, "max_param_abs": 1.978184700012207, "policy_loss": -0.007935994719161954, "value_loss": 3.0093892216682434}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:17:46.437779+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009066.43779, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.9014339143244434, "elapsed_seconds": 6845.162091567996, "env_steps_per_second": 4090.480199802857, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5312344123168233, "eta_seconds": 12712.443884340564, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0030436686431158932, "chaser_grad_norm": 0.5293919295072556, "chaser_new_fields": 54.72500228881836, "chaser_partner_vision": 0.968000054359436, "chaser_return": 14.617500305175781, "chaser_subspace_norm": 94.88986206054688, "collisions_per_episode": 13.175000190734863, "entropy": 0.2534397765994072, "explorer_grad_norm": 0.9321675002574921, "explorer_new_fields": 62.42499923706055, "explorer_partner_vision": 0.968000054359436, "explorer_return": 32.85000228881836, "explorer_subspace_norm": 44.61410903930664, "final_distance": 1.5648239850997925, "max_param_abs": 2.034841299057007, "policy_loss": -0.008697010441707675, "value_loss": 3.0522661209106445}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:25:57.045859+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009557.0458744, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.039756617877768, "elapsed_seconds": 7343.1238243599655, "env_steps_per_second": 4085.4547352828863, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3995943631296135, "eta_seconds": 12238.53970726661, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.00530100038369774, "chaser_grad_norm": 0.48209696263074875, "chaser_new_fields": 56.375, "chaser_partner_vision": 0.9770000576972961, "chaser_return": 16.660001754760742, "chaser_subspace_norm": 114.49951171875, "collisions_per_episode": 14.65000057220459, "entropy": 0.19275250285863876, "explorer_grad_norm": 1.4351917058229446, "explorer_new_fields": 65.20000457763672, "explorer_partner_vision": 0.9770000576972961, "explorer_return": 36.5, "explorer_subspace_norm": 45.261531829833984, "final_distance": 1.5421043634414673, "max_param_abs": 2.0801939964294434, "policy_loss": -0.006963652135809184, "value_loss": 2.7987908124923706}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:34:15.007594+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010055.0076067, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1754005862649906, "elapsed_seconds": 7831.442110553966, "env_steps_per_second": 4086.0928994004203, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2631008793974856, "eta_seconds": 11747.163165830949, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0033816671141939647, "chaser_grad_norm": 0.46328888833522797, "chaser_new_fields": 53.67499923706055, "chaser_partner_vision": 0.9455000162124634, "chaser_return": 15.839999198913574, "chaser_subspace_norm": 111.89578247070312, "collisions_per_episode": 14.350000381469727, "entropy": 0.21325920894742012, "explorer_grad_norm": 0.8308292329311371, "explorer_new_fields": 63.92499923706055, "explorer_partner_vision": 0.9455000162124634, "explorer_return": 34.587501525878906, "explorer_subspace_norm": 46.04932403564453, "final_distance": 1.4381728172302246, "max_param_abs": 2.104952096939087, "policy_loss": -0.006831125341363986, "value_loss": 3.467480182647705}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:23.325878+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010543.3258913, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3100839624991445, "elapsed_seconds": 8316.30226499692, "env_steps_per_second": 4088.3554874027404, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.125407713969431, "eta_seconds": 11251.467770289952, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0050556668809953464, "chaser_grad_norm": 0.6323490589857101, "chaser_new_fields": 59.57500076293945, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 14.142498970031738, "chaser_subspace_norm": 119.39141082763672, "collisions_per_episode": 11.77500057220459, "entropy": 0.1964816302061081, "explorer_grad_norm": 1.1915461868047714, "explorer_new_fields": 66.625, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 39.54999923706055, "explorer_subspace_norm": 51.34266662597656, "final_distance": 1.5876117944717407, "max_param_abs": 2.140456438064575, "policy_loss": -0.007585012739522057, "value_loss": 2.744940936565399}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:28.186032+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011028.1860437, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.444357268364434, "elapsed_seconds": 8799.686166111962, "env_steps_per_second": 4091.0549899651905, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.987547772445419, "eta_seconds": 10755.171980803509, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.003276038641752521, "chaser_grad_norm": 1.419321745634079, "chaser_new_fields": 51.54999923706055, "chaser_partner_vision": 0.971750020980835, "chaser_return": 14.517501831054688, "chaser_subspace_norm": 120.61017608642578, "collisions_per_episode": 13.425000190734863, "entropy": 0.22173313796520233, "explorer_grad_norm": 0.7243275195360184, "explorer_new_fields": 57.22500228881836, "explorer_partner_vision": 0.971750020980835, "explorer_return": 25.375, "explorer_subspace_norm": 51.69886779785156, "final_distance": 1.9679806232452393, "max_param_abs": 2.1565537452697754, "policy_loss": -0.0052130841854989285, "value_loss": 4.078545570373535}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:31.569934+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011511.569949, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5803018565277713, "elapsed_seconds": 9289.086683499976, "env_steps_per_second": 4090.8219822621163, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.851912578267537, "eta_seconds": 10266.885281763132, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.002300439050305414, "chaser_grad_norm": 1.3402238488197327, "chaser_new_fields": 54.525001525878906, "chaser_partner_vision": 0.955500066280365, "chaser_return": 16.707502365112305, "chaser_subspace_norm": 132.30923461914062, "collisions_per_episode": 15.125, "entropy": 0.18779267743229866, "explorer_grad_norm": 1.4113223850727081, "explorer_new_fields": 63.45000076293945, "explorer_partner_vision": 0.955500066280365, "explorer_return": 33.26250076293945, "explorer_subspace_norm": 58.81296157836914, "final_distance": 1.6271885633468628, "max_param_abs": 2.198268175125122, "policy_loss": -0.005038609450252807, "value_loss": 3.4251524209976196}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:40.970453+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012000.9704669, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.718671374670262, "elapsed_seconds": 9787.216948812944, "env_steps_per_second": 4086.9636597614663, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.718671374670262, "eta_seconds": 9787.216948812944, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.00264889863029083, "chaser_grad_norm": 0.6107844263315201, "chaser_new_fields": 62.17499923706055, "chaser_partner_vision": 0.9760000705718994, "chaser_return": 13.144999504089355, "chaser_subspace_norm": 141.69296264648438, "collisions_per_episode": 10.300000190734863, "entropy": 0.18506034836173058, "explorer_grad_norm": 0.8438531458377838, "explorer_new_fields": 70.9749984741211, "explorer_partner_vision": 0.9760000705718994, "explorer_return": 47.9375, "explorer_subspace_norm": 56.29457092285156, "final_distance": 1.7581580877304077, "max_param_abs": 2.221966028213501, "policy_loss": -0.005794895917173193, "value_loss": 2.5580326318740845}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:14:59.100718+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012499.100734, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8541351938346957, "elapsed_seconds": 10274.886697804905, "env_steps_per_second": 4087.636315150098, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.582312794421868, "eta_seconds": 9296.326059918723, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0031877158601094635, "chaser_grad_norm": 0.3025837652385235, "chaser_new_fields": 60.75, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.882501602172852, "chaser_subspace_norm": 148.7667999267578, "collisions_per_episode": 11.375, "entropy": 0.18439796566963196, "explorer_grad_norm": 0.8285828828811646, "explorer_new_fields": 68.3499984741211, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 42.26250076293945, "explorer_subspace_norm": 62.10081100463867, "final_distance": 1.6131728887557983, "max_param_abs": 2.2530040740966797, "policy_loss": -0.006527455376470925, "value_loss": 2.610820174217224}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:06.770465+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012986.7704763, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9882887456549927, "elapsed_seconds": 10757.839484357974, "env_steps_per_second": 4090.04057589598, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.444963519172267, "eta_seconds": 8801.868669020161, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004688273231682416, "chaser_grad_norm": 0.5364589989185333, "chaser_new_fields": 60.07500076293945, "chaser_partner_vision": 0.9752500653266907, "chaser_return": 13.037500381469727, "chaser_subspace_norm": 151.6190948486328, "collisions_per_episode": 10.574999809265137, "entropy": 0.19497263804078102, "explorer_grad_norm": 1.0899477303028107, "explorer_new_fields": 68.1500015258789, "explorer_partner_vision": 0.9752500653266907, "explorer_return": 42.47500228881836, "explorer_subspace_norm": 59.938655853271484, "final_distance": 1.6024208068847656, "max_param_abs": 2.284710168838501, "policy_loss": -0.0059596559871124555, "value_loss": 3.1407867670059204}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:09.723254+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013469.7232711, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1259802501113803, "elapsed_seconds": 11253.528900400968, "env_steps_per_second": 4087.606688277221, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3105071413866725, "eta_seconds": 8317.82570899202, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004601715255304839, "chaser_grad_norm": 0.659913532435894, "chaser_new_fields": 55.625, "chaser_partner_vision": 0.9550000429153442, "chaser_return": 12.1850004196167, "chaser_subspace_norm": 131.7039794921875, "collisions_per_episode": 10.600000381469727, "entropy": 0.21111338213086128, "explorer_grad_norm": 0.703166201710701, "explorer_new_fields": 61.70000076293945, "explorer_partner_vision": 0.9550000429153442, "explorer_return": 34.212501525878906, "explorer_subspace_norm": 60.2715950012207, "final_distance": 1.7229671478271484, "max_param_abs": 2.3001441955566406, "policy_loss": -0.006460366566670928, "value_loss": 3.310504198074341}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:25.412670+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013965.4126842, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2627877994630317, "elapsed_seconds": 11746.036078066914, "env_steps_per_second": 4086.4849793565018, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1751918663086878, "eta_seconds": 7830.6907187112765, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.00374349603211499, "chaser_grad_norm": 0.47442062199115753, "chaser_new_fields": 60.57500076293945, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 9.942500114440918, "chaser_subspace_norm": 143.57444763183594, "collisions_per_episode": 7.625, "entropy": 0.17764540016651154, "explorer_grad_norm": 0.7173541337251663, "explorer_new_fields": 68.0, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 45.22500228881836, "explorer_subspace_norm": 71.3135757446289, "final_distance": 1.707925796508789, "max_param_abs": 2.3741250038146973, "policy_loss": -0.006035578418504883, "value_loss": 2.6776943802833557}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:37.919846+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014457.9198623, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3965089210483064, "elapsed_seconds": 12227.432115773903, "env_steps_per_second": 4089.1660265688893, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.037905352628984, "eta_seconds": 7336.4592694643425, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.004637669390159971, "chaser_grad_norm": 0.7833073437213898, "chaser_new_fields": 55.650001525878906, "chaser_partner_vision": 0.9762500524520874, "chaser_return": 17.725000381469727, "chaser_subspace_norm": 159.64881896972656, "collisions_per_episode": 15.699999809265137, "entropy": 0.21898170188069344, "explorer_grad_norm": 0.8282943069934845, "explorer_new_fields": 62.04999923706055, "explorer_partner_vision": 0.9762500524520874, "explorer_return": 31.774999618530273, "explorer_subspace_norm": 67.76110076904297, "final_distance": 1.709387183189392, "max_param_abs": 2.4182240962982178, "policy_loss": -0.005661805972458467, "value_loss": 3.7847108840942383}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:55:39.315883+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014939.3158944, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5322482356655462, "elapsed_seconds": 12716.093648395967, "env_steps_per_second": 4089.3061531171866, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.901979819204525, "eta_seconds": 6847.12734913629, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.005122404976045658, "chaser_grad_norm": 0.5404690504074097, "chaser_new_fields": 51.45000076293945, "chaser_partner_vision": 0.9495000243186951, "chaser_return": 20.337499618530273, "chaser_subspace_norm": 179.74717712402344, "collisions_per_episode": 18.774999618530273, "entropy": 0.15835684910416603, "explorer_grad_norm": 0.9143335670232773, "explorer_new_fields": 60.5, "explorer_partner_vision": 0.9495000243186951, "explorer_return": 27.274999618530273, "explorer_subspace_norm": 82.8676528930664, "final_distance": 1.6131728887557983, "max_param_abs": 2.4207990169525146, "policy_loss": -0.003387971257780009, "value_loss": 4.1205960512161255}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:03:47.977417+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015427.9774318, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6685900925110846, "elapsed_seconds": 13206.924333039904, "env_steps_per_second": 4088.7642450489116, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.766358192690522, "eta_seconds": 6358.88949368588, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.015038300501669433, "chaser_grad_norm": 0.5875879302620888, "chaser_new_fields": 54.025001525878906, "chaser_partner_vision": 0.9472500681877136, "chaser_return": 12.662501335144043, "chaser_subspace_norm": 166.31260681152344, "collisions_per_episode": 11.52500057220459, "entropy": 0.1527380496263504, "explorer_grad_norm": 1.1693783849477768, "explorer_new_fields": 62.57500076293945, "explorer_partner_vision": 0.9472500681877136, "explorer_return": 34.79999923706055, "explorer_subspace_norm": 82.93914031982422, "final_distance": 1.5938421487808228, "max_param_abs": 2.4432289600372314, "policy_loss": -0.004890929952722445, "value_loss": 3.135464370250702}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:11:58.808101+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015918.8081126, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.8029958702966624, "elapsed_seconds": 13690.785133067984, "env_steps_per_second": 4090.342478952549, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6298553729842837, "eta_seconds": 5867.4793427434215, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0027088768679517927, "chaser_grad_norm": 0.5266599878668785, "chaser_new_fields": 60.72500228881836, "chaser_partner_vision": 0.9670000672340393, "chaser_return": 12.142500877380371, "chaser_subspace_norm": 192.67559814453125, "collisions_per_episode": 9.574999809265137, "entropy": 0.14571769535541534, "explorer_grad_norm": 0.8169879913330078, "explorer_new_fields": 69.5250015258789, "explorer_partner_vision": 0.9670000672340393, "explorer_return": 46.20000076293945, "explorer_subspace_norm": 80.44278717041016, "final_distance": 1.642462134361267, "max_param_abs": 2.491410970687866, "policy_loss": -0.006536863781519742, "value_loss": 2.866920590400696}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:02.668904+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016402.6689203, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9381019337452665, "elapsed_seconds": 14177.166961482959, "env_steps_per_second": 4091.0853457236203, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.493762802455101, "eta_seconds": 5377.546088838364, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.004193384799192845, "chaser_grad_norm": 0.31596722826361656, "chaser_new_fields": 57.900001525878906, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 11.732501029968262, "chaser_subspace_norm": 179.06655883789062, "collisions_per_episode": 9.77500057220459, "entropy": 0.1655190773308277, "explorer_grad_norm": 0.9114458858966827, "explorer_new_fields": 68.30000305175781, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 43.625, "explorer_subspace_norm": 76.99004364013672, "final_distance": 1.8475704193115234, "max_param_abs": 2.530569553375244, "policy_loss": -0.005543691971347187, "value_loss": 2.8706308007240295}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:09.050730+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016889.0507436, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.072133922088025, "elapsed_seconds": 14659.682119516889, "env_steps_per_second": 4092.8581882495355, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3573779740293415, "eta_seconds": 4886.560706505629, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.002374243521035879, "chaser_grad_norm": 0.27748020365834236, "chaser_new_fields": 61.375, "chaser_partner_vision": 0.9647500514984131, "chaser_return": 11.657500267028809, "chaser_subspace_norm": 181.67864990234375, "collisions_per_episode": 8.975000381469727, "entropy": 0.14595364034175873, "explorer_grad_norm": 1.6453716456890106, "explorer_new_fields": 69.92500305175781, "explorer_partner_vision": 0.9647500514984131, "explorer_return": 47.8125, "explorer_subspace_norm": 82.78719329833984, "final_distance": 1.7317100763320923, "max_param_abs": 2.5783498287200928, "policy_loss": -0.0038004378322504806, "value_loss": 2.815015196800232}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:11.565887+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017371.5659037, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.208465974969699, "elapsed_seconds": 15150.477509890916, "env_steps_per_second": 4092.280257141968, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2218127024105578, "eta_seconds": 4398.525728678008, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0035750436370811656, "chaser_grad_norm": 1.0316120982170105, "chaser_new_fields": 57.0, "chaser_partner_vision": 0.9545000195503235, "chaser_return": 12.26500129699707, "chaser_subspace_norm": 181.2757568359375, "collisions_per_episode": 10.449999809265137, "entropy": 0.186406958848238, "explorer_grad_norm": 1.0583333373069763, "explorer_new_fields": 67.6500015258789, "explorer_partner_vision": 0.9545000195503235, "explorer_return": 41.26250076293945, "explorer_subspace_norm": 80.00265502929688, "final_distance": 1.8627468347549438, "max_param_abs": 2.5818021297454834, "policy_loss": -0.0046094126459195195, "value_loss": 3.2537341117858887}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:22.361277+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017862.3612893, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.345760014316636, "elapsed_seconds": 15644.73605153989, "env_steps_per_second": 4090.8328391836667, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.086440003579159, "eta_seconds": 3911.1840128849726, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.004406582113499202, "chaser_grad_norm": 0.42754947394132614, "chaser_new_fields": 57.5, "chaser_partner_vision": 0.9672500491142273, "chaser_return": 14.425000190734863, "chaser_subspace_norm": 184.77886962890625, "collisions_per_episode": 12.449999809265137, "entropy": 0.144505362957716, "explorer_grad_norm": 0.8771203309297562, "explorer_new_fields": 66.55000305175781, "explorer_partner_vision": 0.9672500491142273, "explorer_return": 39.212501525878906, "explorer_subspace_norm": 91.4606704711914, "final_distance": 1.6706188917160034, "max_param_abs": 2.632824659347534, "policy_loss": -0.00546145379172458, "value_loss": 2.7275306582450867}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:52:36.619819+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018356.619835, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.483429296433589, "elapsed_seconds": 16140.345467160922, "env_steps_per_second": 4089.1318054054864, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9510304568192463, "eta_seconds": 3423.7096445492866, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0038937525081870206, "chaser_grad_norm": 0.8020946830511093, "chaser_new_fields": 54.25, "chaser_partner_vision": 0.9652500748634338, "chaser_return": 12.880000114440918, "chaser_subspace_norm": 194.848388671875, "collisions_per_episode": 11.65000057220459, "entropy": 0.16065635159611702, "explorer_grad_norm": 0.7273015528917313, "explorer_new_fields": 63.60000228881836, "explorer_partner_vision": 0.9652500748634338, "explorer_return": 36.53750228881836, "explorer_subspace_norm": 76.82682800292969, "final_distance": 1.682106852531433, "max_param_abs": 2.7014949321746826, "policy_loss": -0.0066974234283065925, "value_loss": 3.14750999212265}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:00:52.229236+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018852.2292495, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.619311195935556, "elapsed_seconds": 16629.520305368, "env_steps_per_second": 4089.1137417866244, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8151725639886276, "eta_seconds": 2934.621230359059, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0030361690627547233, "chaser_grad_norm": 0.5369230136275291, "chaser_new_fields": 61.25, "chaser_partner_vision": 0.9665000438690186, "chaser_return": 12.547499656677246, "chaser_subspace_norm": 216.032470703125, "collisions_per_episode": 10.02500057220459, "entropy": 0.13131140172481537, "explorer_grad_norm": 0.8748937100172043, "explorer_new_fields": 70.0250015258789, "explorer_partner_vision": 0.9665000438690186, "explorer_return": 47.25, "explorer_subspace_norm": 89.8182373046875, "final_distance": 1.4663976430892944, "max_param_abs": 2.7317287921905518, "policy_loss": -0.005932329951775728, "value_loss": 2.2110677361488342}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:09:01.404073+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019341.4040883, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.755061208304154, "elapsed_seconds": 17118.220349894953, "env_steps_per_second": 4089.210126356947, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6792944583291647, "eta_seconds": 2445.460049984993, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0028010765739274746, "chaser_grad_norm": 0.3767981752753258, "chaser_new_fields": 61.875, "chaser_partner_vision": 0.9655000567436218, "chaser_return": 10.324999809265137, "chaser_subspace_norm": 208.04513549804688, "collisions_per_episode": 7.75, "entropy": 0.13262581080198288, "explorer_grad_norm": 0.9192542135715485, "explorer_new_fields": 71.8499984741211, "explorer_partner_vision": 0.9655000567436218, "explorer_return": 51.01250076293945, "explorer_subspace_norm": 91.59584045410156, "final_distance": 1.5939102172851562, "max_param_abs": 2.763204336166382, "policy_loss": -0.004654687465119878, "value_loss": 2.236839532852173}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:17:10.104118+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019830.1041298, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.88903864675693, "elapsed_seconds": 17600.53912832495, "env_steps_per_second": 4090.78378082912, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5432265163063256, "eta_seconds": 1955.615458702772, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0026141230935889936, "chaser_grad_norm": 0.9471372216939926, "chaser_new_fields": 58.92499923706055, "chaser_partner_vision": 0.9615000486373901, "chaser_return": 11.882500648498535, "chaser_subspace_norm": 212.00845336914062, "collisions_per_episode": 9.77500057220459, "entropy": 0.15928909182548523, "explorer_grad_norm": 1.1343470215797424, "explorer_new_fields": 69.125, "explorer_partner_vision": 0.9615000486373901, "explorer_return": 45.38750076293945, "explorer_subspace_norm": 86.06438446044922, "final_distance": 1.7796329259872437, "max_param_abs": 2.813157081604004, "policy_loss": -0.005861258640877054, "value_loss": 2.6857704520225525}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:25:12.422898+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020312.422914, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.025413070107752, "elapsed_seconds": 18091.487052387907, "env_steps_per_second": 4090.321585269172, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4074659246033312, "eta_seconds": 1466.8773285719924, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.006974181910143429, "chaser_grad_norm": 0.5899177044630051, "chaser_new_fields": 59.150001525878906, "chaser_partner_vision": 0.9750000238418579, "chaser_return": 14.177499771118164, "chaser_subspace_norm": 235.09718322753906, "collisions_per_episode": 11.925000190734863, "entropy": 0.14187957346439362, "explorer_grad_norm": 0.7830052524805069, "explorer_new_fields": 67.4000015258789, "explorer_partner_vision": 0.9750000238418579, "explorer_return": 41.462501525878906, "explorer_subspace_norm": 86.65560150146484, "final_distance": 1.6251405477523804, "max_param_abs": 2.8463218212127686, "policy_loss": -0.004497030382992762, "value_loss": 3.0623390078544617}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:33:23.370822+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020803.3708365, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.161473250793028, "elapsed_seconds": 18581.3037028549, "env_steps_per_second": 4090.132813895242, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2716564868838436, "eta_seconds": 977.9633527818369, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.003674934932682561, "chaser_grad_norm": 0.5017296448349953, "chaser_new_fields": 60.125, "chaser_partner_vision": 0.9722500443458557, "chaser_return": 13.889999389648438, "chaser_subspace_norm": 245.86741638183594, "collisions_per_episode": 11.449999809265137, "entropy": 0.14775100722908974, "explorer_grad_norm": 0.9065325260162354, "explorer_new_fields": 70.5999984741211, "explorer_partner_vision": 0.9722500443458557, "explorer_return": 46.23749923706055, "explorer_subspace_norm": 94.84130859375, "final_distance": 1.839495301246643, "max_param_abs": 2.8727216720581055, "policy_loss": -0.0034798192204874567, "value_loss": 2.8518959879875183}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:41:33.187471+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021293.1874876, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.298176957899414, "elapsed_seconds": 19073.43704843789, "env_steps_per_second": 4089.456965827152, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1358506912281901, "eta_seconds": 489.0624884214843, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0029817369262694626, "chaser_grad_norm": 0.3839939646422863, "chaser_new_fields": 59.0, "chaser_partner_vision": 0.971500039100647, "chaser_return": 12.225000381469727, "chaser_subspace_norm": 224.42881774902344, "collisions_per_episode": 10.100000381469727, "entropy": 0.1570109687745571, "explorer_grad_norm": 0.761499211192131, "explorer_new_fields": 69.9749984741211, "explorer_partner_vision": 0.971500039100647, "explorer_return": 46.23749923706055, "explorer_subspace_norm": 95.07978057861328, "final_distance": 1.6962867975234985, "max_param_abs": 2.9067318439483643, "policy_loss": -0.0053824065839762625, "value_loss": 2.6650503873825073}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:49:45.320815+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021785.3208268, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.432533738381107, "elapsed_seconds": 19557.121458171983, "env_steps_per_second": 4090.5815393692224, "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-transformer/social/seed_0003/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:49.005228+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022269.005242, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.432549700080546, "elapsed_seconds": 19557.178920289967, "env_steps_per_second": 4090.569520586759, "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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.002856620680569577, "chaser_grad_norm": 0.7219458222389221, "chaser_new_fields": 60.275001525878906, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 13.242500305175781, "chaser_subspace_norm": 225.597900390625, "collisions_per_episode": 10.875, "entropy": 0.1406782865524292, "explorer_grad_norm": 0.60902339220047, "explorer_new_fields": 70.80000305175781, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 47.42499923706055, "explorer_subspace_norm": 101.78258514404297, "final_distance": 1.8629791736602783, "max_param_abs": 2.9420251846313477, "policy_loss": -0.005349479220825515, "value_loss": 2.428272843360901}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:57:49.062689+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022269.0627043, "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..9b93cd340a9e7a745f250de3c798aa14ef84bbfd --- /dev/null +++ b/social/seed_0003/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.432549700080546, + "elapsed_seconds": 19557.178920289967, + "env_steps_per_second": 4090.569520586759, + "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-transformer/social/seed_0003/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.002856620680569577, + "chaser_grad_norm": 0.7219458222389221, + "chaser_new_fields": 60.275001525878906, + "chaser_partner_vision": 0.9662500619888306, + "chaser_return": 13.242500305175781, + "chaser_subspace_norm": 225.597900390625, + "collisions_per_episode": 10.875, + "entropy": 0.1406782865524292, + "explorer_grad_norm": 0.60902339220047, + "explorer_new_fields": 70.80000305175781, + "explorer_partner_vision": 0.9662500619888306, + "explorer_return": 47.42499923706055, + "explorer_subspace_norm": 101.78258514404297, + "final_distance": 1.8629791736602783, + "max_param_abs": 2.9420251846313477, + "policy_loss": -0.005349479220825515, + "value_loss": 2.428272843360901 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:57:49.062689+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022269.0627043, + "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..16ecb037932f4e7eeaee2828179f148a0639f822 --- /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:b8b54f8a46f42267d8050df70d059bed5fffe88df8f02d91f48faba250143759 +size 29760376 diff --git a/social/seed_0004/attempt_01/config.json b/social/seed_0004/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..8be98dec82511a83a12401e6ccc91585e4a0cbdd --- /dev/null +++ b/social/seed_0004/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0004/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..c51e7bc03367a1773acf17f8bb7f6e0ce6a03140 --- /dev/null +++ b/social/seed_0004/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006006417083295269, "elapsed_seconds": 2.162310149986297, "env_steps_per_second": 1849.8733865839502, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 12.012233524882209, "eta_seconds": 43244.04068957595, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.006405812781538955, "chaser_grad_norm": 3.680071346461773, "chaser_new_fields": 35.72500228881836, "chaser_partner_vision": 0.3615000247955322, "chaser_return": -0.9400001764297485, "chaser_subspace_norm": 0.031355779618024826, "collisions_per_episode": 1.8000000715255737, "entropy": 2.7473949790000916, "explorer_grad_norm": 7.541321873664856, "explorer_new_fields": 31.225000381469727, "explorer_partner_vision": 0.3615000247955322, "explorer_return": -4.4375, "explorer_subspace_norm": 0.03989025577902794, "final_distance": 5.113106727600098, "max_param_abs": 1.001200795173645, "policy_loss": -0.010431485993787959, "value_loss": 7.7662393450737}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:31:54.030598+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783002714.0306137, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13523385432251314, "elapsed_seconds": 486.84187556104735, "env_steps_per_second": 4108.110046398855, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.274120318578013, "eta_seconds": 18986.833146880846, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.003908314893194209, "chaser_grad_norm": 0.7957494705915451, "chaser_new_fields": 55.22500228881836, "chaser_partner_vision": 0.49150002002716064, "chaser_return": 12.692500114440918, "chaser_subspace_norm": 10.50583553314209, "collisions_per_episode": 10.925000190734863, "entropy": 0.7542333751916885, "explorer_grad_norm": 2.036599785089493, "explorer_new_fields": 61.150001525878906, "explorer_partner_vision": 0.49150002002716064, "explorer_return": 33.11249923706055, "explorer_subspace_norm": 4.679419040679932, "final_distance": 5.053833961486816, "max_param_abs": 1.177061676979065, "policy_loss": -0.007037343373518468, "value_loss": 3.5872541666030884}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:39:58.710163+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003198.7101753, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2694732123880789, "elapsed_seconds": 970.1035645970842, "env_steps_per_second": 4123.271108339171, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.1199910353735, "eta_seconds": 18431.9677273446, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0035976978684413297, "chaser_grad_norm": 0.8698631823062897, "chaser_new_fields": 50.82500076293945, "chaser_partner_vision": 0.5727500319480896, "chaser_return": 15.220001220703125, "chaser_subspace_norm": 16.186717987060547, "collisions_per_episode": 14.100000381469727, "entropy": 0.580939993262291, "explorer_grad_norm": 1.3754579722881317, "explorer_new_fields": 63.525001525878906, "explorer_partner_vision": 0.5727500319480896, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 9.646955490112305, "final_distance": 3.6670098304748535, "max_param_abs": 1.306240439414978, "policy_loss": -0.0065794658530471395, "value_loss": 3.7893291115760803}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:48:01.971850+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783003681.971863, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.40721357481806386, "elapsed_seconds": 1465.9688693450298, "env_steps_per_second": 4092.8563528642317, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.022300756089454, "eta_seconds": 18080.282721922034, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004423351254184027, "chaser_grad_norm": 1.9464105367660522, "chaser_new_fields": 47.45000076293945, "chaser_partner_vision": 0.6392500400543213, "chaser_return": 11.315001487731934, "chaser_subspace_norm": 23.454540252685547, "collisions_per_episode": 11.100000381469727, "entropy": 0.4960159510374069, "explorer_grad_norm": 1.707905113697052, "explorer_new_fields": 63.54999923706055, "explorer_partner_vision": 0.6392500400543213, "explorer_return": 36.17499923706055, "explorer_subspace_norm": 14.043740272521973, "final_distance": 3.0441784858703613, "max_param_abs": 1.4073479175567627, "policy_loss": -0.006589006682865772, "value_loss": 3.6471452116966248}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T14:56:17.837156+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004177.8371713, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5447449007002999, "elapsed_seconds": 1961.0816425210796, "env_steps_per_second": 4079.3814120433835, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.902704106302699, "eta_seconds": 17649.734782689717, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.004575278901029378, "chaser_grad_norm": 1.411739632487297, "chaser_new_fields": 46.650001525878906, "chaser_partner_vision": 0.6300000548362732, "chaser_return": 10.30250072479248, "chaser_subspace_norm": 30.85883140563965, "collisions_per_episode": 10.375, "entropy": 0.3799746036529541, "explorer_grad_norm": 2.8338339030742645, "explorer_new_fields": 65.6500015258789, "explorer_partner_vision": 0.6300000548362732, "explorer_return": 38.900001525878906, "explorer_subspace_norm": 21.285064697265625, "final_distance": 3.450026750564575, "max_param_abs": 1.5406363010406494, "policy_loss": -0.008237178688858648, "value_loss": 3.3164199590682983}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:04:32.949931+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783004672.949946, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6817712776302748, "elapsed_seconds": 2454.3765994689893, "env_steps_per_second": 4074.3543603551, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.772398943411924, "eta_seconds": 17180.636196282925, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0023997042723333806, "chaser_grad_norm": 1.2041969746351242, "chaser_new_fields": 49.82500076293945, "chaser_partner_vision": 0.6692500114440918, "chaser_return": 10.892500877380371, "chaser_subspace_norm": 33.56122970581055, "collisions_per_episode": 10.27500057220459, "entropy": 0.36388760060071945, "explorer_grad_norm": 1.7903993725776672, "explorer_new_fields": 65.25, "explorer_partner_vision": 0.6692500114440918, "explorer_return": 39.66250228881836, "explorer_subspace_norm": 23.650028228759766, "final_distance": 2.2592880725860596, "max_param_abs": 1.5879969596862793, "policy_loss": -0.006047059978357083, "value_loss": 3.553192913532257}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:12:46.244885+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005166.2448964, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8223558062858258, "elapsed_seconds": 2960.480902628973, "env_steps_per_second": 4053.395510622525, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.66001623561968, "eta_seconds": 16776.058448230848, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004686810682301257, "chaser_grad_norm": 1.421747013926506, "chaser_new_fields": 47.650001525878906, "chaser_partner_vision": 0.7677500247955322, "chaser_return": 16.69499969482422, "chaser_subspace_norm": 39.68783950805664, "collisions_per_episode": 16.0, "entropy": 0.340407058596611, "explorer_grad_norm": 1.1202066093683243, "explorer_new_fields": 62.275001525878906, "explorer_partner_vision": 0.7677500247955322, "explorer_return": 32.0, "explorer_subspace_norm": 28.11762237548828, "final_distance": 3.26615834236145, "max_param_abs": 1.6506539583206177, "policy_loss": -0.003894635872441654, "value_loss": 3.6380419731140137}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:21:12.349189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783005672.3492048, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9558861938591711, "elapsed_seconds": 3441.1902978930157, "env_steps_per_second": 4068.359720929113, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.506320628193235, "eta_seconds": 16222.754261495646, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.0031292916285234895, "chaser_grad_norm": 1.4354316890239716, "chaser_new_fields": 51.72500228881836, "chaser_partner_vision": 0.9132500290870667, "chaser_return": 17.725000381469727, "chaser_subspace_norm": 46.83580017089844, "collisions_per_episode": 16.350000381469727, "entropy": 0.3061699718236923, "explorer_grad_norm": 1.3145405799150467, "explorer_new_fields": 62.650001525878906, "explorer_partner_vision": 0.9132500290870667, "explorer_return": 31.11250114440918, "explorer_subspace_norm": 30.313024520874023, "final_distance": 2.225423574447632, "max_param_abs": 1.6796084642410278, "policy_loss": -0.007148117741579085, "value_loss": 3.631281852722168}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:29:13.058583+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006153.0585945, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0906811372114074, "elapsed_seconds": 3926.4520939610666, "env_steps_per_second": 4074.925560560946, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.362724548845629, "eta_seconds": 15705.808375844266, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.002195938319509594, "chaser_grad_norm": 0.8357527852058411, "chaser_new_fields": 51.07500076293945, "chaser_partner_vision": 0.9530000686645508, "chaser_return": 20.130001068115234, "chaser_subspace_norm": 51.90518569946289, "collisions_per_episode": 18.850000381469727, "entropy": 0.2868952006101608, "explorer_grad_norm": 1.3366508483886719, "explorer_new_fields": 61.22500228881836, "explorer_partner_vision": 0.9530000686645508, "explorer_return": 25.4375, "explorer_subspace_norm": 31.53109359741211, "final_distance": 1.6400114297866821, "max_param_abs": 1.6956520080566406, "policy_loss": -0.007223165430133349, "value_loss": 3.9085472226142883}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:37:18.320381+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783006638.320396, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2271098139230485, "elapsed_seconds": 4417.595330122975, "env_steps_per_second": 4074.614955620873, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.2267115812905, "eta_seconds": 15216.161692645801, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.009292324467181423, "chaser_grad_norm": 0.858633890748024, "chaser_new_fields": 52.650001525878906, "chaser_partner_vision": 0.9570000171661377, "chaser_return": 21.0575008392334, "chaser_subspace_norm": 62.415321350097656, "collisions_per_episode": 19.375, "entropy": 0.21401458606123924, "explorer_grad_norm": 1.3784401416778564, "explorer_new_fields": 61.525001525878906, "explorer_partner_vision": 0.9570000171661377, "explorer_return": 24.537500381469727, "explorer_subspace_norm": 39.32377243041992, "final_distance": 1.884469985961914, "max_param_abs": 1.7127368450164795, "policy_loss": -0.007042567620942464, "value_loss": 3.490797996520996}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:45:29.463615+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007129.463627, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3640762603188825, "elapsed_seconds": 4910.674537147977, "env_steps_per_second": 4072.760238681101, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.092228780956648, "eta_seconds": 14732.023611443932, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.005526535733622495, "chaser_grad_norm": 0.5741921290755272, "chaser_new_fields": 49.47500228881836, "chaser_partner_vision": 0.9662500619888306, "chaser_return": 18.985000610351562, "chaser_subspace_norm": 63.74802780151367, "collisions_per_episode": 18.05000114440918, "entropy": 0.2565273717045784, "explorer_grad_norm": 1.3801442384719849, "explorer_new_fields": 57.92499923706055, "explorer_partner_vision": 0.9662500619888306, "explorer_return": 20.8125, "explorer_subspace_norm": 39.20549392700195, "final_distance": 1.7350071668624878, "max_param_abs": 1.724376916885376, "policy_loss": -0.009112593937672386, "value_loss": 3.5845817923545837}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T15:53:42.542825+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783007622.5428407, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4980269792391725, "elapsed_seconds": 5392.897125261021, "env_steps_per_second": 4079.439953888455, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.949343854357818, "eta_seconds": 14217.637875688146, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0051266794378683755, "chaser_grad_norm": 0.8771094679832458, "chaser_new_fields": 50.82500076293945, "chaser_partner_vision": 0.9655000567436218, "chaser_return": 20.577499389648438, "chaser_subspace_norm": 71.44198608398438, "collisions_per_episode": 19.375, "entropy": 0.2603532299399376, "explorer_grad_norm": 1.1190015971660614, "explorer_new_fields": 59.20000076293945, "explorer_partner_vision": 0.9655000567436218, "explorer_return": 22.55000114440918, "explorer_subspace_norm": 36.379966735839844, "final_distance": 1.7777214050292969, "max_param_abs": 1.7229582071304321, "policy_loss": -0.0063028686437860415, "value_loss": 3.989771544933319}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:01:44.765411+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008104.7654233, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.636212250518632, "elapsed_seconds": 5890.364101867075, "env_steps_per_second": 4074.4510160912964, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.8178285845434745, "eta_seconds": 13744.182904356509, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.007450404680760414, "chaser_grad_norm": 0.6702429056167603, "chaser_new_fields": 46.125, "chaser_partner_vision": 0.9657500386238098, "chaser_return": 18.880001068115234, "chaser_subspace_norm": 70.05862426757812, "collisions_per_episode": 18.399999618530273, "entropy": 0.24746601656079292, "explorer_grad_norm": 1.4973100125789642, "explorer_new_fields": 53.10000228881836, "explorer_partner_vision": 0.9657500386238098, "explorer_return": 16.625, "explorer_subspace_norm": 48.2133674621582, "final_distance": 1.6684454679489136, "max_param_abs": 1.7499350309371948, "policy_loss": -0.004991331277409472, "value_loss": 4.392935276031494}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:10:02.232387+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783008602.232402, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7710617715258397, "elapsed_seconds": 6375.822377493023, "env_steps_per_second": 4077.9053211678734, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.678359063938282, "eta_seconds": 13242.092630177816, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.006235246734097366, "chaser_grad_norm": 0.5759230926632881, "chaser_new_fields": 51.92499923706055, "chaser_partner_vision": 0.9645000696182251, "chaser_return": 18.020002365112305, "chaser_subspace_norm": 83.27088928222656, "collisions_per_episode": 16.700000762939453, "entropy": 0.23615305870771408, "explorer_grad_norm": 1.1084135919809341, "explorer_new_fields": 60.400001525878906, "explorer_partner_vision": 0.9645000696182251, "explorer_return": 26.325000762939453, "explorer_subspace_norm": 40.690826416015625, "final_distance": 1.584934949874878, "max_param_abs": 1.768629550933838, "policy_loss": -0.007102407678964084, "value_loss": 3.463648557662964}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:18:07.690663+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009087.6906738, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.908880041791126, "elapsed_seconds": 6871.968150448054, "env_steps_per_second": 4074.524122783426, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5450629347549487, "eta_seconds": 12762.226565117815, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0013082725181803312, "chaser_grad_norm": 2.0697110891342163, "chaser_new_fields": 48.70000076293945, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 16.85500144958496, "chaser_subspace_norm": 88.81626892089844, "collisions_per_episode": 16.149999618530273, "entropy": 0.22713254019618034, "explorer_grad_norm": 1.8747320473194122, "explorer_new_fields": 55.47500228881836, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 20.30000114440918, "explorer_subspace_norm": 38.84838104248047, "final_distance": 1.5568745136260986, "max_param_abs": 1.8108711242675781, "policy_loss": -0.0025563403423518594, "value_loss": 4.041815519332886}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:26:23.836436+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783009583.8364513, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.042979691673342, "elapsed_seconds": 7354.726890024031, "env_steps_per_second": 4079.0093838415773, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.4049661527889032, "eta_seconds": 12257.878150040051, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.003859238509242857, "chaser_grad_norm": 1.2495602071285248, "chaser_new_fields": 45.20000076293945, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 19.364999771118164, "chaser_subspace_norm": 96.7882308959961, "collisions_per_episode": 19.149999618530273, "entropy": 0.27206335216760635, "explorer_grad_norm": 0.7675902545452118, "explorer_new_fields": 51.5, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 12.3125, "explorer_subspace_norm": 40.35666275024414, "final_distance": 1.5353316068649292, "max_param_abs": 1.8405003547668457, "policy_loss": -0.005395314401013707, "value_loss": 4.241807699203491}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:34:26.595176+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010066.5951867, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1779492593411125, "elapsed_seconds": 7840.617333628004, "env_steps_per_second": 4081.3112843492113, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2669238890116685, "eta_seconds": 11760.926000442007, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.003307036283928255, "chaser_grad_norm": 0.8185561001300812, "chaser_new_fields": 52.275001525878906, "chaser_partner_vision": 0.9517500400543213, "chaser_return": 12.000000953674316, "chaser_subspace_norm": 97.90601348876953, "collisions_per_episode": 11.050000190734863, "entropy": 0.22399704530835152, "explorer_grad_norm": 1.7540854215621948, "explorer_new_fields": 64.1500015258789, "explorer_partner_vision": 0.9517500400543213, "explorer_return": 37.10000228881836, "explorer_subspace_norm": 42.17123031616211, "final_distance": 1.8330901861190796, "max_param_abs": 1.8503997325897217, "policy_loss": -0.005607030292964055, "value_loss": 3.3499656915664673}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:42:32.485620+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783010552.4856343, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.312399938304734, "elapsed_seconds": 8324.639777897042, "env_steps_per_second": 4084.260809731881, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.128541093000522, "eta_seconds": 11262.74793480188, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0032161123297180794, "chaser_grad_norm": 0.6028102785348892, "chaser_new_fields": 56.25, "chaser_partner_vision": 0.9550000429153442, "chaser_return": 14.837501525878906, "chaser_subspace_norm": 109.61213684082031, "collisions_per_episode": 13.125, "entropy": 0.1798524670302868, "explorer_grad_norm": 1.5265830755233765, "explorer_new_fields": 66.7249984741211, "explorer_partner_vision": 0.9550000429153442, "explorer_return": 37.900001525878906, "explorer_subspace_norm": 46.44622039794922, "final_distance": 1.5440744161605835, "max_param_abs": 1.8859515190124512, "policy_loss": -0.006124499208283396, "value_loss": 2.7894481420516968}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:50:36.508063+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011036.5080748, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.447370208885566, "elapsed_seconds": 8810.532751988037, "env_steps_per_second": 4086.0185204892227, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9912302553045804, "eta_seconds": 10768.42891909649, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0037650154568957817, "chaser_grad_norm": 0.6599537953734398, "chaser_new_fields": 56.92499923706055, "chaser_partner_vision": 0.9742500185966492, "chaser_return": 13.267500877380371, "chaser_subspace_norm": 119.36541748046875, "collisions_per_episode": 11.375, "entropy": 0.18118510767817497, "explorer_grad_norm": 1.214105412364006, "explorer_new_fields": 65.7249984741211, "explorer_partner_vision": 0.9742500185966492, "explorer_return": 38.36249923706055, "explorer_subspace_norm": 50.33469772338867, "final_distance": 1.4625438451766968, "max_param_abs": 1.9162756204605103, "policy_loss": -0.005831173127518241, "value_loss": 2.883459448814392}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T16:58:42.401040+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783011522.4010553, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.583222002317521, "elapsed_seconds": 9299.599208343076, "env_steps_per_second": 4086.1976036460305, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.855140107824629, "eta_seconds": 10278.504388168663, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0039999301656621356, "chaser_grad_norm": 0.4054645262658596, "chaser_new_fields": 55.45000076293945, "chaser_partner_vision": 0.9665000438690186, "chaser_return": 14.355002403259277, "chaser_subspace_norm": 117.07202911376953, "collisions_per_episode": 12.5, "entropy": 0.16044118255376816, "explorer_grad_norm": 1.269740641117096, "explorer_new_fields": 64.7249984741211, "explorer_partner_vision": 0.9665000438690186, "explorer_return": 34.6875, "explorer_subspace_norm": 49.706512451171875, "final_distance": 1.6482959985733032, "max_param_abs": 1.9349397420883179, "policy_loss": -0.006002170505922777, "value_loss": 2.5191983580589294}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:06:51.467494+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012011.467506, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7196398911797184, "elapsed_seconds": 9790.703608246986, "env_steps_per_second": 4085.508212740387, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7196398911797184, "eta_seconds": 9790.703608246986, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0027526107514290743, "chaser_grad_norm": 1.20570769906044, "chaser_new_fields": 55.25, "chaser_partner_vision": 0.9675000309944153, "chaser_return": 14.197501182556152, "chaser_subspace_norm": 122.30333709716797, "collisions_per_episode": 12.675000190734863, "entropy": 0.19474298134446144, "explorer_grad_norm": 1.0700080543756485, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.9675000309944153, "explorer_return": 35.01250076293945, "explorer_subspace_norm": 55.02927017211914, "final_distance": 1.546519160270691, "max_param_abs": 1.9644824266433716, "policy_loss": -0.0067692911801842115, "value_loss": 3.0321977138519287}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:15:02.571894+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012502.5719097, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8577446513025078, "elapsed_seconds": 10287.880744689028, "env_steps_per_second": 4082.4734502955725, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5855784940356026, "eta_seconds": 9308.082578528169, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0037994283712965427, "chaser_grad_norm": 0.5777713656425476, "chaser_new_fields": 57.47500228881836, "chaser_partner_vision": 0.9637500643730164, "chaser_return": 13.592501640319824, "chaser_subspace_norm": 129.4323272705078, "collisions_per_episode": 11.574999809265137, "entropy": 0.1772046722471714, "explorer_grad_norm": 1.2036336064338684, "explorer_new_fields": 65.5, "explorer_partner_vision": 0.9637500643730164, "explorer_return": 38.07500076293945, "explorer_subspace_norm": 55.4228515625, "final_distance": 1.6336511373519897, "max_param_abs": 1.9923532009124756, "policy_loss": -0.005952734657540759, "value_loss": 2.959691643714905}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:23:19.749030+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783012999.7490404, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.994481065504725, "elapsed_seconds": 10780.13183581701, "env_steps_per_second": 4081.5827366609665, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.450029962685684, "eta_seconds": 8820.107865668462, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.005142124885708199, "chaser_grad_norm": 0.8215018659830093, "chaser_new_fields": 52.275001525878906, "chaser_partner_vision": 0.9647500514984131, "chaser_return": 12.780000686645508, "chaser_subspace_norm": 124.1966552734375, "collisions_per_episode": 11.800000190734863, "entropy": 0.19554506614804268, "explorer_grad_norm": 1.0094137787818909, "explorer_new_fields": 60.67499923706055, "explorer_partner_vision": 0.9647500514984131, "explorer_return": 30.950000762939453, "explorer_subspace_norm": 65.90809631347656, "final_distance": 1.6938421726226807, "max_param_abs": 2.0822455883026123, "policy_loss": -0.006410844996580067, "value_loss": 3.5163609385490417}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:31:32.000122+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013492.0001376, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.130436664388625, "elapsed_seconds": 11269.57199179905, "env_steps_per_second": 4081.787669795671, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3138010128089834, "eta_seconds": 8329.68364611234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0020310347875029944, "chaser_grad_norm": 0.5331714823842049, "chaser_new_fields": 56.625, "chaser_partner_vision": 0.9782500267028809, "chaser_return": 14.410000801086426, "chaser_subspace_norm": 115.28544616699219, "collisions_per_episode": 12.550000190734863, "entropy": 0.1890837326645851, "explorer_grad_norm": 0.9485040903091431, "explorer_new_fields": 62.650001525878906, "explorer_partner_vision": 0.9782500267028809, "explorer_return": 33.91250228881836, "explorer_subspace_norm": 64.95726776123047, "final_distance": 1.6542656421661377, "max_param_abs": 2.1536245346069336, "policy_loss": -0.004516581768641004, "value_loss": 3.2359978556632996}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:39:41.440277+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783013981.4402878, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2677835035911023, "elapsed_seconds": 11764.020612927969, "env_steps_per_second": 4080.237665280084, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1785223357274015, "eta_seconds": 7842.680408618646, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0020988428769044276, "chaser_grad_norm": 0.9987809360027313, "chaser_new_fields": 53.57500076293945, "chaser_partner_vision": 0.971250057220459, "chaser_return": 14.157501220703125, "chaser_subspace_norm": 156.90481567382812, "collisions_per_episode": 12.77500057220459, "entropy": 0.17717525362968445, "explorer_grad_norm": 1.2092021703720093, "explorer_new_fields": 61.5, "explorer_partner_vision": 0.971250057220459, "explorer_return": 31.475000381469727, "explorer_subspace_norm": 63.91344451904297, "final_distance": 1.662379503250122, "max_param_abs": 2.149197816848755, "policy_loss": -0.005861490251490409, "value_loss": 3.2724303007125854}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:47:55.888900+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014475.8889153, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.4028413822916788, "elapsed_seconds": 12250.228976250044, "env_steps_per_second": 4081.556360859604, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0417048293750075, "eta_seconds": 7350.137385750027, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0031276542944311814, "chaser_grad_norm": 0.9992583841085434, "chaser_new_fields": 55.04999923706055, "chaser_partner_vision": 0.9475000500679016, "chaser_return": 12.414999961853027, "chaser_subspace_norm": 153.40989685058594, "collisions_per_episode": 10.949999809265137, "entropy": 0.20114708319306374, "explorer_grad_norm": 1.2479537576436996, "explorer_new_fields": 63.525001525878906, "explorer_partner_vision": 0.9475000500679016, "explorer_return": 36.10000228881836, "explorer_subspace_norm": 61.7996940612793, "final_distance": 1.7645915746688843, "max_param_abs": 2.1441197395324707, "policy_loss": -0.005308174765659146, "value_loss": 3.0697078108787537}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T17:56:02.097262+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783014962.0972724, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.537920290173885, "elapsed_seconds": 12736.513044625986, "env_steps_per_second": 4082.7501073334006, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.9050340024013228, "eta_seconds": 6858.122408644762, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.004685206818905385, "chaser_grad_norm": 0.9260435700416565, "chaser_new_fields": 57.04999923706055, "chaser_partner_vision": 0.971250057220459, "chaser_return": 15.954998970031738, "chaser_subspace_norm": 165.13096618652344, "collisions_per_episode": 13.949999809265137, "entropy": 0.17736686021089554, "explorer_grad_norm": 1.7992304861545563, "explorer_new_fields": 67.45000457763672, "explorer_partner_vision": 0.971250057220459, "explorer_return": 39.51250076293945, "explorer_subspace_norm": 70.3255615234375, "final_distance": 1.7520240545272827, "max_param_abs": 2.15751314163208, "policy_loss": -0.0047795680137028285, "value_loss": 3.19892817735672}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:04:08.381331+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015448.3813455, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6754283086238946, "elapsed_seconds": 13231.54191104602, "env_steps_per_second": 4081.157008233444, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7696506671152086, "eta_seconds": 6370.742401614751, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.003842664841338106, "chaser_grad_norm": 0.4684142470359802, "chaser_new_fields": 57.32500076293945, "chaser_partner_vision": 0.9790000319480896, "chaser_return": 13.482500076293945, "chaser_subspace_norm": 159.06610107421875, "collisions_per_episode": 11.675000190734863, "entropy": 0.15418144315481186, "explorer_grad_norm": 1.49385267496109, "explorer_new_fields": 65.92500305175781, "explorer_partner_vision": 0.9790000319480896, "explorer_return": 38.70000076293945, "explorer_subspace_norm": 72.39019012451172, "final_distance": 1.6523526906967163, "max_param_abs": 2.1608173847198486, "policy_loss": -0.005249279722576716, "value_loss": 2.907705068588257}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:12:23.410197+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783015943.4102066, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.812326779806689, "elapsed_seconds": 13724.37640730408, "env_steps_per_second": 4080.3311085374294, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6338543342028666, "eta_seconds": 5881.87560313032, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.003635764474800851, "chaser_grad_norm": 1.1646097004413605, "chaser_new_fields": 51.97500228881836, "chaser_partner_vision": 0.971000075340271, "chaser_return": 16.97749900817871, "chaser_subspace_norm": 174.71722412109375, "collisions_per_episode": 15.925000190734863, "entropy": 0.15988433733582497, "explorer_grad_norm": 1.303291916847229, "explorer_new_fields": 59.650001525878906, "explorer_partner_vision": 0.971000075340271, "explorer_return": 26.225000381469727, "explorer_subspace_norm": 74.96487426757812, "final_distance": 1.671054482460022, "max_param_abs": 2.191709518432617, "policy_loss": -0.004008417759109761, "value_loss": 3.261890947818756}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:20:36.244695+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016436.2447116, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.948258116921384, "elapsed_seconds": 14213.729220916983, "env_steps_per_second": 4080.5617652154906, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4976151477977664, "eta_seconds": 5391.414532071959, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0025776667835324396, "chaser_grad_norm": 0.633307583630085, "chaser_new_fields": 56.47500228881836, "chaser_partner_vision": 0.9727500677108765, "chaser_return": 14.067500114440918, "chaser_subspace_norm": 174.60317993164062, "collisions_per_episode": 12.225000381469727, "entropy": 0.2069396749138832, "explorer_grad_norm": 1.1711878329515457, "explorer_new_fields": 63.82500076293945, "explorer_partner_vision": 0.9727500677108765, "explorer_return": 35.61249923706055, "explorer_subspace_norm": 69.92669677734375, "final_distance": 1.7697423696517944, "max_param_abs": 2.2128281593322754, "policy_loss": -0.004609539440821209, "value_loss": 3.28468656539917}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:28:45.597506+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783016925.597518, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.084072249666949, "elapsed_seconds": 14702.660098801018, "env_steps_per_second": 4080.8941781149465, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.36135741655565, "eta_seconds": 4900.88669960034, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.009004256984185943, "chaser_grad_norm": 0.7878830283880234, "chaser_new_fields": 47.45000076293945, "chaser_partner_vision": 0.9602500200271606, "chaser_return": 24.774999618530273, "chaser_subspace_norm": 154.4032745361328, "collisions_per_episode": 23.600000381469727, "entropy": 0.24312667921185493, "explorer_grad_norm": 1.1869306117296219, "explorer_new_fields": 55.97500228881836, "explorer_partner_vision": 0.9602500200271606, "explorer_return": 17.92500114440918, "explorer_subspace_norm": 71.80714416503906, "final_distance": 1.6882448196411133, "max_param_abs": 2.2400803565979004, "policy_loss": -0.005212186070195912, "value_loss": 4.87359619140625}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:36:54.528385+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017414.5283988, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.218807086558631, "elapsed_seconds": 15187.705511611071, "env_steps_per_second": 4082.2492872673038, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2248149606137961, "eta_seconds": 4409.333858209666, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0044931633121336745, "chaser_grad_norm": 0.9938799887895584, "chaser_new_fields": 53.54999923706055, "chaser_partner_vision": 0.9302500486373901, "chaser_return": 16.19500160217285, "chaser_subspace_norm": 185.31402587890625, "collisions_per_episode": 14.75, "entropy": 0.20310849323868752, "explorer_grad_norm": 0.7461930587887764, "explorer_new_fields": 60.525001525878906, "explorer_partner_vision": 0.9302500486373901, "explorer_return": 29.36250114440918, "explorer_subspace_norm": 77.26957702636719, "final_distance": 1.704933524131775, "max_param_abs": 2.256072521209717, "policy_loss": -0.005260136698325368, "value_loss": 3.6548393964767456}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:44:59.573796+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783017899.5738075, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.360229056506683, "elapsed_seconds": 15696.824603424058, "env_steps_per_second": 4077.2577649902028, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0900572641266708, "eta_seconds": 3924.2061508560146, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0031913102403109406, "chaser_grad_norm": 0.9235102757811546, "chaser_new_fields": 50.70000076293945, "chaser_partner_vision": 0.9510000348091125, "chaser_return": 14.010001182556152, "chaser_subspace_norm": 191.9345703125, "collisions_per_episode": 13.25, "entropy": 0.21017852798104286, "explorer_grad_norm": 0.6502271592617035, "explorer_new_fields": 57.25, "explorer_partner_vision": 0.9510000348091125, "explorer_return": 25.162500381469727, "explorer_subspace_norm": 73.7095947265625, "final_distance": 1.6229671239852905, "max_param_abs": 2.2582414150238037, "policy_loss": -0.0039548941613247734, "value_loss": 3.635454475879669}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T18:53:28.692890+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018408.6929052, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.497825730485007, "elapsed_seconds": 16192.172629746026, "env_steps_per_second": 4076.0435001016417, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9540842458604561, "eta_seconds": 3434.703285097642, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0026333769233630733, "chaser_grad_norm": 0.2897908501327038, "chaser_new_fields": 56.82500076293945, "chaser_partner_vision": 0.9722500443458557, "chaser_return": 13.280000686645508, "chaser_subspace_norm": 197.0770263671875, "collisions_per_episode": 11.449999809265137, "entropy": 0.17452480271458626, "explorer_grad_norm": 1.7367521822452545, "explorer_new_fields": 64.2750015258789, "explorer_partner_vision": 0.9722500443458557, "explorer_return": 36.86249923706055, "explorer_subspace_norm": 76.11566162109375, "final_distance": 1.6371471881866455, "max_param_abs": 2.2995553016662598, "policy_loss": -0.003543555010580235, "value_loss": 2.878424108028412}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:01:44.040916+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783018904.0409284, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.6329717505428, "elapsed_seconds": 16678.69830195408, "env_steps_per_second": 4077.056780386339, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8175832500957882, "eta_seconds": 2943.2997003448377, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.003598232948293756, "chaser_grad_norm": 0.5918027982115746, "chaser_new_fields": 57.04999923706055, "chaser_partner_vision": 0.9665000438690186, "chaser_return": 14.132501602172852, "chaser_subspace_norm": 212.28909301757812, "collisions_per_episode": 12.27500057220459, "entropy": 0.17651503905653954, "explorer_grad_norm": 1.1177958995103836, "explorer_new_fields": 62.70000076293945, "explorer_partner_vision": 0.9665000438690186, "explorer_return": 34.04999923706055, "explorer_subspace_norm": 82.35449981689453, "final_distance": 1.586163878440857, "max_param_abs": 2.3393642902374268, "policy_loss": -0.005307407168743339, "value_loss": 2.9786505103111267}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:09:50.566588+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019390.5666041, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.7701771702783295, "elapsed_seconds": 17172.637813001988, "env_steps_per_second": 4076.252044808202, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6814538814683329, "eta_seconds": 2453.233973285998, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.005341162024581989, "chaser_grad_norm": 1.1465555429458618, "chaser_new_fields": 54.32500076293945, "chaser_partner_vision": 0.9590000510215759, "chaser_return": 15.360000610351562, "chaser_subspace_norm": 213.8933563232422, "collisions_per_episode": 13.75, "entropy": 0.20223139598965645, "explorer_grad_norm": 1.2481859475374222, "explorer_new_fields": 62.32500076293945, "explorer_partner_vision": 0.9590000510215759, "explorer_return": 32.07500076293945, "explorer_subspace_norm": 87.7821044921875, "final_distance": 1.681313395500183, "max_param_abs": 2.37076735496521, "policy_loss": -0.002816241706710798, "value_loss": 3.235797643661499}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:18:04.506098+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783019884.5061104, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.908256888723052, "elapsed_seconds": 17669.72479940299, "env_steps_per_second": 4074.7663485077414, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5453618765247836, "eta_seconds": 1963.302755489221, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.002983825173330956, "chaser_grad_norm": 1.2265490889549255, "chaser_new_fields": 61.29999923706055, "chaser_partner_vision": 0.9722500443458557, "chaser_return": 11.642500877380371, "chaser_subspace_norm": 220.87936401367188, "collisions_per_episode": 9.02500057220459, "entropy": 0.16817110404372215, "explorer_grad_norm": 1.4352666586637497, "explorer_new_fields": 66.75, "explorer_partner_vision": 0.9722500443458557, "explorer_return": 42.72500228881836, "explorer_subspace_norm": 85.05738067626953, "final_distance": 1.5028175115585327, "max_param_abs": 2.410098075866699, "policy_loss": -0.004764340856127092, "value_loss": 2.7416463494300842}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:26:21.593086+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020381.5930996, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.04204577842973, "elapsed_seconds": 18151.36480234703, "env_steps_per_second": 4076.828426170552, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.40881452257538353, "eta_seconds": 1471.7322812713808, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.002253646903337625, "chaser_grad_norm": 0.36236999183893204, "chaser_new_fields": 57.625, "chaser_partner_vision": 0.9792500734329224, "chaser_return": 12.40250015258789, "chaser_subspace_norm": 240.12225341796875, "collisions_per_episode": 10.574999809265137, "entropy": 0.16590595990419388, "explorer_grad_norm": 0.712123915553093, "explorer_new_fields": 64.80000305175781, "explorer_partner_vision": 0.9792500734329224, "explorer_return": 37.60000228881836, "explorer_subspace_norm": 86.13605499267578, "final_distance": 1.6343870162963867, "max_param_abs": 2.453023910522461, "policy_loss": -0.004853722718536346, "value_loss": 2.993720829486847}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:34:23.233087+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783020863.2331007, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.176713101345005, "elapsed_seconds": 18636.167164842016, "env_steps_per_second": 4078.091773257834, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27245858428131603, "eta_seconds": 980.8509034127377, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0035411540801248975, "chaser_grad_norm": 0.33849726989865303, "chaser_new_fields": 59.79999923706055, "chaser_partner_vision": 0.9435000419616699, "chaser_return": 11.242500305175781, "chaser_subspace_norm": 230.09683227539062, "collisions_per_episode": 9.02500057220459, "entropy": 0.1792609617114067, "explorer_grad_norm": 0.9835126549005508, "explorer_new_fields": 66.4000015258789, "explorer_partner_vision": 0.9435000419616699, "explorer_return": 42.462501525878906, "explorer_subspace_norm": 87.55894470214844, "final_distance": 1.8121521472930908, "max_param_abs": 2.496182680130005, "policy_loss": -0.0055387865946840975, "value_loss": 3.0503124594688416}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:42:28.035451+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021348.0354688, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.313663486286953, "elapsed_seconds": 19129.18855063303, "env_steps_per_second": 4077.5383542036757, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13624778169966545, "eta_seconds": 490.4920141187956, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.004099105874082909, "chaser_grad_norm": 0.5882994458079338, "chaser_new_fields": 59.275001525878906, "chaser_partner_vision": 0.9832500219345093, "chaser_return": 12.039999961853027, "chaser_subspace_norm": 232.50941467285156, "collisions_per_episode": 9.90000057220459, "entropy": 0.14788555726408958, "explorer_grad_norm": 1.0904473960399628, "explorer_new_fields": 64.0999984741211, "explorer_partner_vision": 0.9832500219345093, "explorer_return": 38.04999923706055, "explorer_subspace_norm": 92.65904235839844, "final_distance": 1.5467513799667358, "max_param_abs": 2.5170822143554688, "policy_loss": -0.0050339502682845705, "value_loss": 2.6593138575553894}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:50:41.056836+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783021841.0568473, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.44470816855002, "elapsed_seconds": 19600.94940678007, "env_steps_per_second": 4081.4349519379703, "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-transformer/social/seed_0004/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:32.817692+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022312.8177032, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.444820641110548, "elapsed_seconds": 19601.35430799797, "env_steps_per_second": 4081.350642560319, "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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.004475183506503555, "chaser_grad_norm": 0.6177457571029663, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9817500710487366, "chaser_return": 9.820000648498535, "chaser_subspace_norm": 235.4537811279297, "collisions_per_episode": 7.550000190734863, "entropy": 0.16257280483841896, "explorer_grad_norm": 1.1996710449457169, "explorer_new_fields": 68.70000457763672, "explorer_partner_vision": 0.9817500710487366, "explorer_return": 47.10000228881836, "explorer_subspace_norm": 95.60823822021484, "final_distance": 1.5751407146453857, "max_param_abs": 2.528623104095459, "policy_loss": -0.005616524885397212, "value_loss": 2.955764353275299}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-02T19:58:33.222594+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022313.2226105, "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..647e5b6427ec593202eef36ba711a255b3274c89 --- /dev/null +++ b/social/seed_0004/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.444820641110548, + "elapsed_seconds": 19601.35430799797, + "env_steps_per_second": 4081.350642560319, + "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-transformer/social/seed_0004/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.004475183506503555, + "chaser_grad_norm": 0.6177457571029663, + "chaser_new_fields": 59.85000228881836, + "chaser_partner_vision": 0.9817500710487366, + "chaser_return": 9.820000648498535, + "chaser_subspace_norm": 235.4537811279297, + "collisions_per_episode": 7.550000190734863, + "entropy": 0.16257280483841896, + "explorer_grad_norm": 1.1996710449457169, + "explorer_new_fields": 68.70000457763672, + "explorer_partner_vision": 0.9817500710487366, + "explorer_return": 47.10000228881836, + "explorer_subspace_norm": 95.60823822021484, + "final_distance": 1.5751407146453857, + "max_param_abs": 2.528623104095459, + "policy_loss": -0.005616524885397212, + "value_loss": 2.955764353275299 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-02T19:58:33.222594+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783022313.2226105, + "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..2d32c58815d093fa42a2409afbc579982b59d0ed --- /dev/null +++ b/social/seed_0005/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:78eaccb338b45f6a03f515f2cee3766c9898f3ac6218f1e0839ec5175bd738c0 +size 29760344 diff --git a/social/seed_0005/attempt_01/config.json b/social/seed_0005/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..c650498d7bf2455d3e81cf5c5d9d280bdb5da3c4 --- /dev/null +++ b/social/seed_0005/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0005/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..b9447262f689565c99597742308860d06098f43b --- /dev/null +++ b/social/seed_0005/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006982248858225325, "elapsed_seconds": 2.513609588961117, "env_steps_per_second": 1591.3370228879548, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.963799491564828, "eta_seconds": 50269.67816963338, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.005349934594416572, "chaser_grad_norm": 3.4319381415843964, "chaser_new_fields": 34.650001525878906, "chaser_partner_vision": 0.320000022649765, "chaser_return": -1.0450001955032349, "chaser_subspace_norm": 0.033552832901477814, "collisions_per_episode": 1.899999976158142, "entropy": 2.751994550228119, "explorer_grad_norm": 6.695566713809967, "explorer_new_fields": 34.54999923706055, "explorer_partner_vision": 0.320000022649765, "explorer_return": 0.5, "explorer_subspace_norm": 0.039362117648124695, "final_distance": 5.559739589691162, "max_param_abs": 1.0011998414993286, "policy_loss": -0.01049435127619347, "value_loss": 7.886138081550598}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:48.317181+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022268.3171976, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13073197409054652, "elapsed_seconds": 470.6351067259675, "env_steps_per_second": 4249.576734539105, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.0985469895313145, "eta_seconds": 18354.769162312732, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0027531317050888404, "chaser_grad_norm": 1.4015900790691376, "chaser_new_fields": 61.5, "chaser_partner_vision": 0.44575002789497375, "chaser_return": 11.960000991821289, "chaser_subspace_norm": 9.958410263061523, "collisions_per_episode": 9.0, "entropy": 0.6765478849411011, "explorer_grad_norm": 1.5145222544670105, "explorer_new_fields": 66.0999984741211, "explorer_partner_vision": 0.44575002789497375, "explorer_return": 42.25, "explorer_subspace_norm": 6.548254013061523, "final_distance": 4.741616725921631, "max_param_abs": 1.2457081079483032, "policy_loss": -0.005068872299972327, "value_loss": 3.0397000908851624}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:05:36.438676+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022736.4386876, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2670633157947062, "elapsed_seconds": 961.4279368609423, "env_steps_per_second": 4160.478228935162, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.074203000099417, "eta_seconds": 18267.130800357903, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00331529213323023, "chaser_grad_norm": 0.7538688480854034, "chaser_new_fields": 54.275001525878906, "chaser_partner_vision": 0.4820000231266022, "chaser_return": 11.585000038146973, "chaser_subspace_norm": 16.687259674072266, "collisions_per_episode": 9.949999809265137, "entropy": 0.4652421325445175, "explorer_grad_norm": 2.8302724361419678, "explorer_new_fields": 67.70000457763672, "explorer_partner_vision": 0.4820000231266022, "explorer_return": 44.13750076293945, "explorer_subspace_norm": 13.17489242553711, "final_distance": 3.988475799560547, "max_param_abs": 1.393466830253601, "policy_loss": -0.005978314537543383, "value_loss": 3.181817650794983}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:13:47.231505+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023227.2315185, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4011734212088696, "elapsed_seconds": 1444.2243163519306, "env_steps_per_second": 4154.479281415112, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.947805528242725, "eta_seconds": 17812.09990167381, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.004081674360833176, "chaser_grad_norm": 0.8306653201580048, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.5060000419616699, "chaser_return": 9.542500495910645, "chaser_subspace_norm": 20.451068878173828, "collisions_per_episode": 7.525000095367432, "entropy": 0.46000928431749344, "explorer_grad_norm": 1.5927399396896362, "explorer_new_fields": 68.5250015258789, "explorer_partner_vision": 0.5060000419616699, "explorer_return": 46.587501525878906, "explorer_subspace_norm": 17.14387321472168, "final_distance": 3.5119309425354004, "max_param_abs": 1.5482282638549805, "policy_loss": -0.007248810037054909, "value_loss": 2.784811854362488}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:21:50.027883+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023710.0278952, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5377929144049994, "elapsed_seconds": 1936.054491857998, "env_steps_per_second": 4132.115099881584, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.840136229644995, "eta_seconds": 17424.490426721983, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.004838004107721616, "chaser_grad_norm": 1.2997898161411285, "chaser_new_fields": 52.5, "chaser_partner_vision": 0.5532500147819519, "chaser_return": 12.4375, "chaser_subspace_norm": 27.19570541381836, "collisions_per_episode": 11.125, "entropy": 0.40200865268707275, "explorer_grad_norm": 1.2595272064208984, "explorer_new_fields": 67.07500457763672, "explorer_partner_vision": 0.5532500147819519, "explorer_return": 41.9375, "explorer_subspace_norm": 22.007015228271484, "final_distance": 3.304868459701538, "max_param_abs": 1.6445814371109009, "policy_loss": -0.00699144538118901, "value_loss": 3.412033438682556}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:01.858061+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024201.858075, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6726647627413816, "elapsed_seconds": 2421.593145868974, "env_steps_per_second": 4129.512844492117, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.70865333918967, "eta_seconds": 16951.152021082813, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0032668352581874394, "chaser_grad_norm": 1.068993479013443, "chaser_new_fields": 51.625, "chaser_partner_vision": 0.612000048160553, "chaser_return": 12.247501373291016, "chaser_subspace_norm": 36.86787033081055, "collisions_per_episode": 11.225000381469727, "entropy": 0.36411894112825394, "explorer_grad_norm": 1.839378297328949, "explorer_new_fields": 66.5, "explorer_partner_vision": 0.612000048160553, "explorer_return": 40.3125, "explorer_subspace_norm": 25.65043830871582, "final_distance": 3.2272987365722656, "max_param_abs": 1.7576699256896973, "policy_loss": -0.008814245636557416, "value_loss": 3.313090741634369}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:07.396714+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024687.3967288, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8109967827133369, "elapsed_seconds": 2919.5884177680127, "env_steps_per_second": 4110.16838091646, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.595648435375575, "eta_seconds": 16544.334367352072, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.005335831306211958, "chaser_grad_norm": 0.7101929038763046, "chaser_new_fields": 51.60000228881836, "chaser_partner_vision": 0.6360000371932983, "chaser_return": 12.132501602172852, "chaser_subspace_norm": 37.11115264892578, "collisions_per_episode": 11.074999809265137, "entropy": 0.36320243775844574, "explorer_grad_norm": 1.4115231037139893, "explorer_new_fields": 66.375, "explorer_partner_vision": 0.6360000371932983, "explorer_return": 40.837501525878906, "explorer_subspace_norm": 27.983671188354492, "final_distance": 3.31510853767395, "max_param_abs": 1.7973805665969849, "policy_loss": -0.010109068686141143, "value_loss": 3.494069516658783}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:46:25.391986+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025185.3920014, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9447511969838848, "elapsed_seconds": 3401.104309141985, "env_steps_per_second": 4116.310094450428, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.453827071495456, "eta_seconds": 16033.777457383643, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004905105728798986, "chaser_grad_norm": 0.8915857821702957, "chaser_new_fields": 53.32500076293945, "chaser_partner_vision": 0.6325000524520874, "chaser_return": 9.647500991821289, "chaser_subspace_norm": 39.08972930908203, "collisions_per_episode": 8.425000190734863, "entropy": 0.3527367413043976, "explorer_grad_norm": 1.7772577106952667, "explorer_new_fields": 68.20000457763672, "explorer_partner_vision": 0.6325000524520874, "explorer_return": 45.04999923706055, "explorer_subspace_norm": 31.200824737548828, "final_distance": 3.7083327770233154, "max_param_abs": 1.8436424732208252, "policy_loss": -0.008357025785358019, "value_loss": 3.036491811275482}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:54:26.907876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025666.9078884, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0781145642544532, "elapsed_seconds": 3881.212431316031, "env_steps_per_second": 4122.423155945309, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.312458257017813, "eta_seconds": 15524.849725264125, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0060263812858778465, "chaser_grad_norm": 1.0612146705389023, "chaser_new_fields": 49.35000228881836, "chaser_partner_vision": 0.7697500586509705, "chaser_return": 12.905000686645508, "chaser_subspace_norm": 48.0130500793457, "collisions_per_episode": 12.25, "entropy": 0.3474693298339844, "explorer_grad_norm": 1.6651727259159088, "explorer_new_fields": 66.20000457763672, "explorer_partner_vision": 0.7697500586509705, "explorer_return": 38.63750076293945, "explorer_subspace_norm": 30.656423568725586, "final_distance": 2.6213314533233643, "max_param_abs": 1.8503252267837524, "policy_loss": -0.006100784500119971, "value_loss": 3.617547035217285}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:02:27.015999+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026147.016014, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2128707881180323, "elapsed_seconds": 4366.334837224917, "env_steps_per_second": 4122.450675688478, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.177666047962111, "eta_seconds": 15039.597772663601, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.003724661815869723, "chaser_grad_norm": 0.9076564311981201, "chaser_new_fields": 49.20000076293945, "chaser_partner_vision": 0.8815000653266907, "chaser_return": 12.650001525878906, "chaser_subspace_norm": 51.69980239868164, "collisions_per_episode": 12.100000381469727, "entropy": 0.32777274399995804, "explorer_grad_norm": 1.076069250702858, "explorer_new_fields": 62.400001525878906, "explorer_partner_vision": 0.8815000653266907, "explorer_return": 34.0625, "explorer_subspace_norm": 34.355987548828125, "final_distance": 2.6171042919158936, "max_param_abs": 1.8954073190689087, "policy_loss": -0.007951904643391572, "value_loss": 3.5293240547180176}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:10:32.138404+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026632.1384153, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.353655849027758, "elapsed_seconds": 4873.161056499928, "env_steps_per_second": 4104.112252420545, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.060967547083273, "eta_seconds": 14619.483169499783, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0037115655905979317, "chaser_grad_norm": 1.1271285116672516, "chaser_new_fields": 51.42499923706055, "chaser_partner_vision": 0.9305000305175781, "chaser_return": 15.835000991821289, "chaser_subspace_norm": 60.358158111572266, "collisions_per_episode": 14.5, "entropy": 0.2508811354637146, "explorer_grad_norm": 2.2482080459594727, "explorer_new_fields": 62.125, "explorer_partner_vision": 0.9305000305175781, "explorer_return": 30.649999618530273, "explorer_subspace_norm": 36.09610366821289, "final_distance": 1.9809786081314087, "max_param_abs": 1.905994176864624, "policy_loss": -0.00444221126008415, "value_loss": 3.2930103540420532}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:18:58.964624+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027138.9646378, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4882009980066666, "elapsed_seconds": 5357.523592824, "env_steps_per_second": 4106.374823895754, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9234389947448483, "eta_seconds": 14124.380381081453, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.007425447994473533, "chaser_grad_norm": 0.8600679188966751, "chaser_new_fields": 48.10000228881836, "chaser_partner_vision": 0.9225000143051147, "chaser_return": 20.014999389648438, "chaser_subspace_norm": 65.72338104248047, "collisions_per_episode": 19.149999618530273, "entropy": 0.37216032296419144, "explorer_grad_norm": 1.1129421889781952, "explorer_new_fields": 56.375, "explorer_partner_vision": 0.9225000143051147, "explorer_return": 20.86250114440918, "explorer_subspace_norm": 33.53558349609375, "final_distance": 1.9673947095870972, "max_param_abs": 1.9331660270690918, "policy_loss": -0.007317121436869911, "value_loss": 4.7356462478637695}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:03.327160+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027623.3271725, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6230172992769318, "elapsed_seconds": 5842.862277396955, "env_steps_per_second": 4107.575852479653, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7870403649795072, "eta_seconds": 13633.345313926226, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.005413036010356365, "chaser_grad_norm": 1.121791586279869, "chaser_new_fields": 49.70000076293945, "chaser_partner_vision": 0.9430000185966492, "chaser_return": 17.78999900817871, "chaser_subspace_norm": 85.87894439697266, "collisions_per_episode": 16.700000762939453, "entropy": 0.23779676854610443, "explorer_grad_norm": 1.2925776541233063, "explorer_new_fields": 59.025001525878906, "explorer_partner_vision": 0.9430000185966492, "explorer_return": 25.3125, "explorer_subspace_norm": 36.831748962402344, "final_distance": 1.4542654752731323, "max_param_abs": 1.9667357206344604, "policy_loss": -0.005185484075923341, "value_loss": 3.923635184764862}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:08.665844+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028108.665859, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.757789823515276, "elapsed_seconds": 6328.043364654994, "env_steps_per_second": 4108.694979118166, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6507942488394196, "eta_seconds": 13142.85929582191, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.003306688547718817, "chaser_grad_norm": 0.6032348871231079, "chaser_new_fields": 49.20000076293945, "chaser_partner_vision": 0.9590000510215759, "chaser_return": 19.560001373291016, "chaser_subspace_norm": 94.8508071899414, "collisions_per_episode": 18.5, "entropy": 0.20687514543533325, "explorer_grad_norm": 1.228437900543213, "explorer_new_fields": 58.67499923706055, "explorer_partner_vision": 0.9590000510215759, "explorer_return": 23.850000381469727, "explorer_subspace_norm": 44.906288146972656, "final_distance": 1.6521590948104858, "max_param_abs": 2.019134521484375, "policy_loss": -0.004534192409828108, "value_loss": 3.933926820755005}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:13.846933+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028593.8469481, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.896144037697199, "elapsed_seconds": 6826.118535709917, "env_steps_per_second": 4101.891851646259, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.52141035572337, "eta_seconds": 12677.077280604131, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0031356794902315244, "chaser_grad_norm": 0.8597266525030136, "chaser_new_fields": 49.07500076293945, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 17.28999900817871, "chaser_subspace_norm": 83.23550415039062, "collisions_per_episode": 16.5, "entropy": 0.24811818450689316, "explorer_grad_norm": 1.3889343738555908, "explorer_new_fields": 56.5, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 21.887500762939453, "explorer_subspace_norm": 44.21297836303711, "final_distance": 1.7542694807052612, "max_param_abs": 2.0504848957061768, "policy_loss": -0.00686289619838476, "value_loss": 4.129640817642212}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:51:31.922105+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029091.9221208, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.03095500115137, "elapsed_seconds": 7311.438004144933, "env_steps_per_second": 4103.160005322165, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3849250019189503, "eta_seconds": 12185.730006908221, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0022722993817367865, "chaser_grad_norm": 0.5147045254707336, "chaser_new_fields": 54.025001525878906, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 18.33500099182129, "chaser_subspace_norm": 99.50810241699219, "collisions_per_episode": 16.80000114440918, "entropy": 0.19260162115097046, "explorer_grad_norm": 1.2161296606063843, "explorer_new_fields": 58.775001525878906, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 24.587499618530273, "explorer_subspace_norm": 47.25221633911133, "final_distance": 1.708525538444519, "max_param_abs": 2.077687978744507, "policy_loss": -0.004209087148311141, "value_loss": 3.195892095565796}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:59:37.241571+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029577.2415829, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1648490348083174, "elapsed_seconds": 7793.456525309943, "env_steps_per_second": 4106.008662020139, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.247273552212476, "eta_seconds": 11690.184787964914, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.005941147318427653, "chaser_grad_norm": 0.4446476474404335, "chaser_new_fields": 50.79999923706055, "chaser_partner_vision": 0.9495000243186951, "chaser_return": 16.23000144958496, "chaser_subspace_norm": 102.59766387939453, "collisions_per_episode": 15.40000057220459, "entropy": 0.2584521472454071, "explorer_grad_norm": 0.8836365193128586, "explorer_new_fields": 58.150001525878906, "explorer_partner_vision": 0.9495000243186951, "explorer_return": 24.462499618530273, "explorer_subspace_norm": 40.52704620361328, "final_distance": 1.6462215185165405, "max_param_abs": 2.1167914867401123, "policy_loss": -0.006910176267851398, "value_loss": 3.660647451877594}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:07:39.260094+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030059.2601085, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.299739106396931, "elapsed_seconds": 8279.060783028952, "env_steps_per_second": 4106.746029657831, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.111411732184083, "eta_seconds": 11201.0822358627, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.005292822174808975, "chaser_grad_norm": 0.33410024642944336, "chaser_new_fields": 50.150001525878906, "chaser_partner_vision": 0.9542500376701355, "chaser_return": 20.152502059936523, "chaser_subspace_norm": 112.93301391601562, "collisions_per_episode": 19.075000762939453, "entropy": 0.19193671643733978, "explorer_grad_norm": 1.1707675009965897, "explorer_new_fields": 56.47500228881836, "explorer_partner_vision": 0.9542500376701355, "explorer_return": 19.850000381469727, "explorer_subspace_norm": 53.89468765258789, "final_distance": 1.5685135126113892, "max_param_abs": 2.14009165763855, "policy_loss": -0.0045728406570901825, "value_loss": 3.632633924484253}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:15:44.864350+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030544.8643625, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.436946176581389, "elapsed_seconds": 8773.006235693, "env_steps_per_second": 4103.496456383891, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9784897713772525, "eta_seconds": 10722.563176958109, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.004049505086792582, "chaser_grad_norm": 0.7459895834326744, "chaser_new_fields": 47.32500076293945, "chaser_partner_vision": 0.9632500410079956, "chaser_return": 17.045001983642578, "chaser_subspace_norm": 133.1437530517578, "collisions_per_episode": 16.700000762939453, "entropy": 0.1951727792620659, "explorer_grad_norm": 1.2430985569953918, "explorer_new_fields": 54.70000076293945, "explorer_partner_vision": 0.9632500410079956, "explorer_return": 19.875, "explorer_subspace_norm": 52.10408020019531, "final_distance": 1.8251793384552002, "max_param_abs": 2.197279691696167, "policy_loss": -0.005414603115987404, "value_loss": 3.7766984701156616}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:23:58.809804+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031038.8098176, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.573309819715279, "elapsed_seconds": 9263.915350975003, "env_steps_per_second": 4101.937308397426, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8441845375800447, "eta_seconds": 10239.064335288162, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00419992365780919, "chaser_grad_norm": 0.494183711707592, "chaser_new_fields": 48.47500228881836, "chaser_partner_vision": 0.9437500238418579, "chaser_return": 16.502500534057617, "chaser_subspace_norm": 129.5998992919922, "collisions_per_episode": 15.875, "entropy": 0.20660506188869476, "explorer_grad_norm": 0.9218032509088516, "explorer_new_fields": 57.70000076293945, "explorer_partner_vision": 0.9437500238418579, "explorer_return": 23.100000381469727, "explorer_subspace_norm": 48.31619644165039, "final_distance": 1.780974268913269, "max_param_abs": 2.2616822719573975, "policy_loss": -0.00570681838317455, "value_loss": 3.506541132926941}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:09.718918+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031529.7189305, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.7122811994833236, "elapsed_seconds": 9764.212318139966, "env_steps_per_second": 4096.592607443404, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.7122811994833236, "eta_seconds": 9764.212318139966, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.004840586327233609, "chaser_grad_norm": 0.9499641060829163, "chaser_new_fields": 53.275001525878906, "chaser_partner_vision": 0.9540000557899475, "chaser_return": 17.364999771118164, "chaser_subspace_norm": 142.5840606689453, "collisions_per_episode": 15.850000381469727, "entropy": 0.1808273084461689, "explorer_grad_norm": 0.9583985656499863, "explorer_new_fields": 61.95000076293945, "explorer_partner_vision": 0.9540000557899475, "explorer_return": 30.387500762939453, "explorer_subspace_norm": 60.58298873901367, "final_distance": 1.7493470907211304, "max_param_abs": 2.304556131362915, "policy_loss": -0.00648990182578757, "value_loss": 3.679859161376953}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:40:30.015888+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032030.0159032, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8497231626875066, "elapsed_seconds": 10259.003385675023, "env_steps_per_second": 4093.964922425696, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.578320956717268, "eta_seconds": 9281.955444182164, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0036848339096562377, "chaser_grad_norm": 0.7695183008909225, "chaser_new_fields": 56.95000076293945, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 11.40250015258789, "chaser_subspace_norm": 135.3102264404297, "collisions_per_episode": 9.625, "entropy": 0.20266563817858696, "explorer_grad_norm": 1.4999758899211884, "explorer_new_fields": 65.42500305175781, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 39.462501525878906, "explorer_subspace_norm": 56.27037048339844, "final_distance": 1.734866738319397, "max_param_abs": 2.335583448410034, "policy_loss": -0.006669347069234277, "value_loss": 2.7285598516464233}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:48:44.806955+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032524.8069694, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9877569762661005, "elapsed_seconds": 10755.925114557962, "env_steps_per_second": 4090.768532819808, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4445284351268093, "eta_seconds": 8800.302366456513, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0035449450995574394, "chaser_grad_norm": 0.412155594676733, "chaser_new_fields": 53.5, "chaser_partner_vision": 0.9707500338554382, "chaser_return": 14.489999771118164, "chaser_subspace_norm": 144.37506103515625, "collisions_per_episode": 12.949999809265137, "entropy": 0.1804564818739891, "explorer_grad_norm": 1.3178684264421463, "explorer_new_fields": 66.57500457763672, "explorer_partner_vision": 0.9707500338554382, "explorer_return": 38.66250228881836, "explorer_subspace_norm": 60.74799728393555, "final_distance": 1.8221911191940308, "max_param_abs": 2.395131826400757, "policy_loss": -0.0032319089290808733, "value_loss": 2.9465296864509583}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:01.728684+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033021.728698, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.127735217934985, "elapsed_seconds": 11259.846784565947, "env_steps_per_second": 4085.3131379241267, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.311804291517163, "eta_seconds": 8322.495449461787, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.001854412824503271, "chaser_grad_norm": 0.4784124121069908, "chaser_new_fields": 53.150001525878906, "chaser_partner_vision": 0.9732500314712524, "chaser_return": 17.80000114440918, "chaser_subspace_norm": 145.3020477294922, "collisions_per_episode": 16.350000381469727, "entropy": 0.16520023345947266, "explorer_grad_norm": 0.8667920380830765, "explorer_new_fields": 61.875, "explorer_partner_vision": 0.9732500314712524, "explorer_return": 28.375, "explorer_subspace_norm": 67.08818817138672, "final_distance": 1.5876117944717407, "max_param_abs": 2.4005322456359863, "policy_loss": -0.004553586544489008, "value_loss": 2.6065053939819336}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:25.650352+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033525.650364, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2665896900088733, "elapsed_seconds": 11759.722884031944, "env_steps_per_second": 4081.728836074639, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1777264600059154, "eta_seconds": 7839.815256021296, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.005144789809106598, "chaser_grad_norm": 0.6557071506977081, "chaser_new_fields": 53.400001525878906, "chaser_partner_vision": 0.9585000276565552, "chaser_return": 13.702499389648438, "chaser_subspace_norm": 150.7068328857422, "collisions_per_episode": 12.574999809265137, "entropy": 0.19064639136195183, "explorer_grad_norm": 1.2867234498262405, "explorer_new_fields": 62.17499923706055, "explorer_partner_vision": 0.9585000276565552, "explorer_return": 32.587501525878906, "explorer_subspace_norm": 56.772029876708984, "final_distance": 1.7316421270370483, "max_param_abs": 2.413417100906372, "policy_loss": -0.004908868611792627, "value_loss": 2.8435784578323364}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:13:45.526452+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034025.5264673, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.4020654818324774, "elapsed_seconds": 12247.435734596918, "env_steps_per_second": 4082.4872310828723, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0412392890994866, "eta_seconds": 7348.461440758151, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.006624455454270262, "chaser_grad_norm": 0.7127513587474823, "chaser_new_fields": 52.67499923706055, "chaser_partner_vision": 0.956000030040741, "chaser_return": 16.7450008392334, "chaser_subspace_norm": 167.2206573486328, "collisions_per_episode": 15.40000057220459, "entropy": 0.21796854957938194, "explorer_grad_norm": 0.7433254420757294, "explorer_new_fields": 61.10000228881836, "explorer_partner_vision": 0.956000030040741, "explorer_return": 29.86250114440918, "explorer_subspace_norm": 58.1534423828125, "final_distance": 1.5669010877609253, "max_param_abs": 2.452385187149048, "policy_loss": -0.005440361294873242, "value_loss": 3.7050411701202393}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:21:53.239302+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034513.2393143, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.539294879047763, "elapsed_seconds": 12741.461564571946, "env_steps_per_second": 4081.164451697419, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.905774165641103, "eta_seconds": 6860.78699630797, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.005458471066621229, "chaser_grad_norm": 0.8215318024158478, "chaser_new_fields": 56.22500228881836, "chaser_partner_vision": 0.9650000333786011, "chaser_return": 14.847500801086426, "chaser_subspace_norm": 183.37928771972656, "collisions_per_episode": 12.875, "entropy": 0.22322159260511398, "explorer_grad_norm": 0.7849171161651611, "explorer_new_fields": 63.45000076293945, "explorer_partner_vision": 0.9650000333786011, "explorer_return": 34.837501525878906, "explorer_subspace_norm": 64.34459686279297, "final_distance": 1.6971737146377563, "max_param_abs": 2.4753785133361816, "policy_loss": -0.005358163624080703, "value_loss": 3.4562771320343018}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:30:07.265133+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035007.265148, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6728163823433313, "elapsed_seconds": 13222.138976435992, "env_steps_per_second": 4084.0593262736693, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7683930729801225, "eta_seconds": 6366.215062728441, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0023644274090734996, "chaser_grad_norm": 0.42591068893671036, "chaser_new_fields": 56.85000228881836, "chaser_partner_vision": 0.9772500395774841, "chaser_return": 16.8125, "chaser_subspace_norm": 189.13058471679688, "collisions_per_episode": 14.675000190734863, "entropy": 0.20647722482681274, "explorer_grad_norm": 1.4939558804035187, "explorer_new_fields": 63.67499923706055, "explorer_partner_vision": 0.9772500395774841, "explorer_return": 34.20000076293945, "explorer_subspace_norm": 66.93545532226562, "final_distance": 1.8205612897872925, "max_param_abs": 2.492704391479492, "policy_loss": -0.005143799840539032, "value_loss": 2.959179401397705}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:38:07.942543+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035487.9425554, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.810138144221661, "elapsed_seconds": 13716.49731919798, "env_steps_per_second": 4082.6749495019317, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.632916347523569, "eta_seconds": 5878.498851084849, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0039734480803250705, "chaser_grad_norm": 0.533793717622757, "chaser_new_fields": 55.45000076293945, "chaser_partner_vision": 0.9782500267028809, "chaser_return": 16.44500160217285, "chaser_subspace_norm": 188.4586639404297, "collisions_per_episode": 14.550000190734863, "entropy": 0.19092320278286934, "explorer_grad_norm": 0.9036921858787537, "explorer_new_fields": 61.17499923706055, "explorer_partner_vision": 0.9782500267028809, "explorer_return": 30.887500762939453, "explorer_subspace_norm": 66.60169982910156, "final_distance": 1.7638275623321533, "max_param_abs": 2.550511598587036, "policy_loss": -0.0036644012253077562, "value_loss": 3.2592252492904663}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:46:22.300888+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035982.3009028, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.947023508973329, "elapsed_seconds": 14209.284632303985, "env_steps_per_second": 4081.8381432194246, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4971468482312627, "eta_seconds": 5389.7286536325455, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.004551818977800437, "chaser_grad_norm": 0.54115329682827, "chaser_new_fields": 55.150001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 12.945001602172852, "chaser_subspace_norm": 191.050537109375, "collisions_per_episode": 11.449999809265137, "entropy": 0.1809760332107544, "explorer_grad_norm": 0.8418004214763641, "explorer_new_fields": 63.04999923706055, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 36.0, "explorer_subspace_norm": 74.13909149169922, "final_distance": 1.7646595239639282, "max_param_abs": 2.574176549911499, "policy_loss": -0.00439433621426133, "value_loss": 3.193272292613983}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:54:35.088199+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036475.0882113, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.082863395194452, "elapsed_seconds": 14698.308222700027, "env_steps_per_second": 4082.1024495415168, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3609544650648175, "eta_seconds": 4899.436074233343, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.00235088825287022, "chaser_grad_norm": 0.46493207663297653, "chaser_new_fields": 57.70000076293945, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 12.960000991821289, "chaser_subspace_norm": 209.1356658935547, "collisions_per_episode": 10.949999809265137, "entropy": 0.14787404611706734, "explorer_grad_norm": 0.9296358227729797, "explorer_new_fields": 64.20000457763672, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 36.8125, "explorer_subspace_norm": 76.02066040039062, "final_distance": 1.6898921728134155, "max_param_abs": 2.6117730140686035, "policy_loss": -0.00393303696158398, "value_loss": 2.7516806721687317}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:02:44.111791+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036964.1118078, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.216410021253881, "elapsed_seconds": 15179.076076513971, "env_steps_per_second": 4084.5700810427015, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.224119038428546, "eta_seconds": 4406.828538342766, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.005242228043059072, "chaser_grad_norm": 0.3406782038509846, "chaser_new_fields": 60.75, "chaser_partner_vision": 0.956000030040741, "chaser_return": 11.370000839233398, "chaser_subspace_norm": 217.9557342529297, "collisions_per_episode": 8.800000190734863, "entropy": 0.1581752598285675, "explorer_grad_norm": 0.7001714408397675, "explorer_new_fields": 69.45000457763672, "explorer_partner_vision": 0.956000030040741, "explorer_return": 45.72500228881836, "explorer_subspace_norm": 81.10758972167969, "final_distance": 2.0356028079986572, "max_param_abs": 2.64066481590271, "policy_loss": -0.0055387803558346604, "value_loss": 2.659282624721527}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:10:44.879643+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037444.8796558, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.349790389712758, "elapsed_seconds": 15659.245402965927, "env_steps_per_second": 4087.0424054966356, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0874475974281894, "eta_seconds": 3914.811350741482, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.003204891581958691, "chaser_grad_norm": 0.43026255816221237, "chaser_new_fields": 60.125, "chaser_partner_vision": 0.9747500419616699, "chaser_return": 10.342500686645508, "chaser_subspace_norm": 217.27061462402344, "collisions_per_episode": 8.02500057220459, "entropy": 0.16142263263463974, "explorer_grad_norm": 1.1604721993207932, "explorer_new_fields": 63.525001525878906, "explorer_partner_vision": 0.9747500419616699, "explorer_return": 39.17499923706055, "explorer_subspace_norm": 83.94412231445312, "final_distance": 1.878540277481079, "max_param_abs": 2.685065746307373, "policy_loss": -0.005360821680170602, "value_loss": 2.807082951068878}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:18:45.048970+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037925.0489848, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.484701661843897, "elapsed_seconds": 16144.92598263803, "env_steps_per_second": 4087.971668063096, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9513003525123418, "eta_seconds": 3424.6812690444303, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0027848600847342198, "chaser_grad_norm": 0.605685405433178, "chaser_new_fields": 58.20000076293945, "chaser_partner_vision": 0.9622500538825989, "chaser_return": 13.660000801086426, "chaser_subspace_norm": 213.88784790039062, "collisions_per_episode": 11.449999809265137, "entropy": 0.16796811297535896, "explorer_grad_norm": 1.037843570113182, "explorer_new_fields": 65.5250015258789, "explorer_partner_vision": 0.9622500538825989, "explorer_return": 39.22500228881836, "explorer_subspace_norm": 82.40847778320312, "final_distance": 1.518677830696106, "max_param_abs": 2.694932460784912, "policy_loss": -0.004647529352496349, "value_loss": 3.2572144865989685}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:26:50.729550+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038410.7295635, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.620142575363312, "elapsed_seconds": 16632.513271307922, "env_steps_per_second": 4088.3779192471206, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8153192780052902, "eta_seconds": 2935.149400819045, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.005685233005664259, "chaser_grad_norm": 0.8081489652395248, "chaser_new_fields": 52.75, "chaser_partner_vision": 0.9637500643730164, "chaser_return": 15.117502212524414, "chaser_subspace_norm": 195.9572296142578, "collisions_per_episode": 13.77500057220459, "entropy": 0.19451913610100746, "explorer_grad_norm": 0.7126703411340714, "explorer_new_fields": 58.47500228881836, "explorer_partner_vision": 0.9637500643730164, "explorer_return": 27.787500381469727, "explorer_subspace_norm": 83.29158020019531, "final_distance": 1.6669009923934937, "max_param_abs": 2.74478816986084, "policy_loss": -0.004704660574930131, "value_loss": 4.004489064216614}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:58.316838+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038898.3168535, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.753874657714979, "elapsed_seconds": 17113.948767773923, "env_steps_per_second": 4090.2307789895976, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6791249511021398, "eta_seconds": 2444.8498239677033, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0025955632756975433, "chaser_grad_norm": 0.49055852741003036, "chaser_new_fields": 59.0, "chaser_partner_vision": 0.9792500734329224, "chaser_return": 9.947501182556152, "chaser_subspace_norm": 217.71546936035156, "collisions_per_episode": 7.775000095367432, "entropy": 0.1540343277156353, "explorer_grad_norm": 0.7066687494516373, "explorer_new_fields": 66.5999984741211, "explorer_partner_vision": 0.9792500734329224, "explorer_return": 43.01250076293945, "explorer_subspace_norm": 91.65914154052734, "final_distance": 1.76891028881073, "max_param_abs": 2.784733295440674, "policy_loss": -0.005424137379065996, "value_loss": 2.6861993074417114}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:59.752336+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039379.7523475, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.888180577850823, "elapsed_seconds": 17597.450080262963, "env_steps_per_second": 4091.5018750787153, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.543131175316758, "eta_seconds": 1955.272231140329, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.004019093479315394, "chaser_grad_norm": 0.4551091641187668, "chaser_new_fields": 60.67499923706055, "chaser_partner_vision": 0.9630000591278076, "chaser_return": 11.28499984741211, "chaser_subspace_norm": 256.4726257324219, "collisions_per_episode": 8.75, "entropy": 0.1491725742816925, "explorer_grad_norm": 0.9486653059720993, "explorer_new_fields": 68.4749984741211, "explorer_partner_vision": 0.9630000591278076, "explorer_return": 45.337501525878906, "explorer_subspace_norm": 85.718994140625, "final_distance": 1.8254116773605347, "max_param_abs": 2.823303461074829, "policy_loss": -0.00370861927976307, "value_loss": 2.708367347717285}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:51:03.253650+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039863.253665, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.022869436620256, "elapsed_seconds": 18082.32997183292, "env_steps_per_second": 4092.3929667952502, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.407259684050291, "eta_seconds": 1466.1348625810476, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.004851702444817363, "chaser_grad_norm": 0.4657657891511917, "chaser_new_fields": 55.72500228881836, "chaser_partner_vision": 0.9595000743865967, "chaser_return": 13.432501792907715, "chaser_subspace_norm": 236.9140167236328, "collisions_per_episode": 11.824999809265137, "entropy": 0.14498943835496902, "explorer_grad_norm": 1.1784703582525253, "explorer_new_fields": 63.025001525878906, "explorer_partner_vision": 0.9595000743865967, "explorer_return": 34.91250228881836, "explorer_subspace_norm": 94.32679748535156, "final_distance": 1.8188421726226807, "max_param_abs": 2.823406457901001, "policy_loss": -0.004510773004085422, "value_loss": 3.20566064119339}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:59:08.133539+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040348.1335518, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.159152019428891, "elapsed_seconds": 18572.94726994401, "env_steps_per_second": 4091.9730668157504, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2715343168120469, "eta_seconds": 977.5235405233689, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0024754665879704874, "chaser_grad_norm": 0.34231916815042496, "chaser_new_fields": 58.525001525878906, "chaser_partner_vision": 0.9600000381469727, "chaser_return": 9.860000610351562, "chaser_subspace_norm": 236.02944946289062, "collisions_per_episode": 7.849999904632568, "entropy": 0.15843476355075836, "explorer_grad_norm": 1.010767176747322, "explorer_new_fields": 67.0999984741211, "explorer_partner_vision": 0.9600000381469727, "explorer_return": 43.16250228881836, "explorer_subspace_norm": 94.40425872802734, "final_distance": 1.8099727630615234, "max_param_abs": 2.8220231533050537, "policy_loss": -0.0047451577312358495, "value_loss": 2.793255627155304}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:07:18.750837+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040838.7508523, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.295683427349432, "elapsed_seconds": 19064.460338457953, "env_steps_per_second": 4091.3825314348815, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1357867545474213, "eta_seconds": 488.83231637071674, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.002585947495195029, "chaser_grad_norm": 0.7114167511463165, "chaser_new_fields": 56.70000076293945, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 15.547500610351562, "chaser_subspace_norm": 198.20169067382812, "collisions_per_episode": 13.475000381469727, "entropy": 0.1851428598165512, "explorer_grad_norm": 2.5365684032440186, "explorer_new_fields": 65.6500015258789, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 37.6875, "explorer_subspace_norm": 84.80267333984375, "final_distance": 1.9968732595443726, "max_param_abs": 2.850515604019165, "policy_loss": -0.005180863614199893, "value_loss": 3.449484944343567}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:15:30.263905+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041330.2639186, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.425836457755811, "elapsed_seconds": 19533.01124792092, "env_steps_per_second": 4095.630672844421, "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-transformer/social/seed_0005/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:18.814815+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041798.8148248, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.4258453104955455, "elapsed_seconds": 19533.043117783964, "env_steps_per_second": 4095.623990465857, "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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.003451461161873226, "chaser_grad_norm": 0.4533055052161217, "chaser_new_fields": 55.75, "chaser_partner_vision": 0.9477500319480896, "chaser_return": 9.160000801086426, "chaser_subspace_norm": 194.28756713867188, "collisions_per_episode": 7.650000095367432, "entropy": 0.20668550580739975, "explorer_grad_norm": 0.7168894857168198, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9477500319480896, "explorer_return": 39.0625, "explorer_subspace_norm": 89.7675552368164, "final_distance": 1.5833224058151245, "max_param_abs": 2.8546435832977295, "policy_loss": -0.007046535808569843, "value_loss": 3.2317103147506714}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:18.846686+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041798.8466992, "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..dfe907d6af1f4c1e36a64dabbb07c37ee09c68fc --- /dev/null +++ b/social/seed_0005/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.4258453104955455, + "elapsed_seconds": 19533.043117783964, + "env_steps_per_second": 4095.623990465857, + "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-transformer/social/seed_0005/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.003451461161873226, + "chaser_grad_norm": 0.4533055052161217, + "chaser_new_fields": 55.75, + "chaser_partner_vision": 0.9477500319480896, + "chaser_return": 9.160000801086426, + "chaser_subspace_norm": 194.28756713867188, + "collisions_per_episode": 7.650000095367432, + "entropy": 0.20668550580739975, + "explorer_grad_norm": 0.7168894857168198, + "explorer_new_fields": 63.625, + "explorer_partner_vision": 0.9477500319480896, + "explorer_return": 39.0625, + "explorer_subspace_norm": 89.7675552368164, + "final_distance": 1.5833224058151245, + "max_param_abs": 2.8546435832977295, + "policy_loss": -0.007046535808569843, + "value_loss": 3.2317103147506714 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:18.846686+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041798.8466992, + "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..4777bc359cf27f66a9fbe2b78d4091427722657b --- /dev/null +++ b/social/seed_0006/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:23e8a61977d52ecaa6a6bca53b80271a5b31024659be373ca295584a0cf5e3fa +size 29760368 diff --git a/social/seed_0006/attempt_01/config.json b/social/seed_0006/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..94a355e69839c99f7496fd3c651d23cd5b778df3 --- /dev/null +++ b/social/seed_0006/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0006/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..f29314d7c9d9b1a96d159dc63bce1eb8d273810c --- /dev/null +++ b/social/seed_0006/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006733056486377285, "elapsed_seconds": 2.423900335095823, "env_steps_per_second": 1650.232867285721, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 13.465439667105933, "eta_seconds": 48475.58280158136, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.006195690087260175, "chaser_grad_norm": 6.170320704579353, "chaser_new_fields": 34.54999923706055, "chaser_partner_vision": 0.36125001311302185, "chaser_return": -1.5075002908706665, "chaser_subspace_norm": 0.03308657184243202, "collisions_per_episode": 1.475000023841858, "entropy": 2.7509084343910217, "explorer_grad_norm": 6.949445605278015, "explorer_new_fields": 33.625, "explorer_partner_vision": 0.36125001311302185, "explorer_return": -0.6000000238418579, "explorer_subspace_norm": 0.033542171120643616, "final_distance": 5.313870906829834, "max_param_abs": 1.0011979341506958, "policy_loss": -0.012727415928637242, "value_loss": 7.689297318458557}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:57:59.117813+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022279.117826, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13193274965027943, "elapsed_seconds": 474.9578987410059, "env_steps_per_second": 4210.899545626039, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.145377236360898, "eta_seconds": 18523.35805089923, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.010044509555314107, "chaser_grad_norm": 0.6857345700263977, "chaser_new_fields": 55.35000228881836, "chaser_partner_vision": 0.4897500276565552, "chaser_return": 13.239999771118164, "chaser_subspace_norm": 8.581465721130371, "collisions_per_episode": 11.350000381469727, "entropy": 0.7754151970148087, "explorer_grad_norm": 2.5638015270233154, "explorer_new_fields": 62.35000228881836, "explorer_partner_vision": 0.4897500276565552, "explorer_return": 35.11249923706055, "explorer_subspace_norm": 4.941499710083008, "final_distance": 4.44049596786499, "max_param_abs": 1.2035341262817383, "policy_loss": -0.005265899775547922, "value_loss": 3.7983439564704895}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:05:51.651812+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022751.6518245, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2653431866836359, "elapsed_seconds": 955.2354720610892, "env_steps_per_second": 4187.44918608319, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.041520546989082, "eta_seconds": 18149.473969160696, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.005884241392171141, "chaser_grad_norm": 0.767892599105835, "chaser_new_fields": 53.95000076293945, "chaser_partner_vision": 0.5515000224113464, "chaser_return": 13.527501106262207, "chaser_subspace_norm": 17.8148193359375, "collisions_per_episode": 11.875, "entropy": 0.5374384671449661, "explorer_grad_norm": 2.5021607279777527, "explorer_new_fields": 62.025001525878906, "explorer_partner_vision": 0.5515000224113464, "explorer_return": 33.76250076293945, "explorer_subspace_norm": 8.91495132446289, "final_distance": 4.082591533660889, "max_param_abs": 1.305163025856018, "policy_loss": -0.006032783747642156, "value_loss": 3.8397581577301025}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:13:51.929385+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023231.9293993, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.400166063533074, "elapsed_seconds": 1440.5978287190665, "env_steps_per_second": 4164.93755605269, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.935381450241246, "eta_seconds": 17767.373220868485, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.003966091229671065, "chaser_grad_norm": 0.6331578344106674, "chaser_new_fields": 45.125, "chaser_partner_vision": 0.643750011920929, "chaser_return": 15.332501411437988, "chaser_subspace_norm": 23.615434646606445, "collisions_per_episode": 15.225000381469727, "entropy": 0.4996975436806679, "explorer_grad_norm": 2.7831935584545135, "explorer_new_fields": 59.025001525878906, "explorer_partner_vision": 0.643750011920929, "explorer_return": 27.512500762939453, "explorer_subspace_norm": 12.837922096252441, "final_distance": 3.557846784591675, "max_param_abs": 1.4353307485580444, "policy_loss": -0.004926918209534392, "value_loss": 4.327607989311218}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:21:57.291740+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023717.2917535, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5350563473344664, "elapsed_seconds": 1926.202850404079, "env_steps_per_second": 4153.248967688818, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.815507126010198, "eta_seconds": 17335.82565363671, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0028720776767041656, "chaser_grad_norm": 1.648443877696991, "chaser_new_fields": 45.07500076293945, "chaser_partner_vision": 0.6492500305175781, "chaser_return": 12.032501220703125, "chaser_subspace_norm": 35.482826232910156, "collisions_per_episode": 12.27500057220459, "entropy": 0.3810296133160591, "explorer_grad_norm": 3.621328115463257, "explorer_new_fields": 63.85000228881836, "explorer_partner_vision": 0.6492500305175781, "explorer_return": 35.212501525878906, "explorer_subspace_norm": 17.693885803222656, "final_distance": 3.0730180740356445, "max_param_abs": 1.508556604385376, "policy_loss": -0.005074748618994818, "value_loss": 3.4802892208099365}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:02.896766+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024202.89678, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6700119061725046, "elapsed_seconds": 2412.0428622210165, "env_steps_per_second": 4145.863308080674, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.690083343207532, "eta_seconds": 16884.300035547116, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.003359872362973526, "chaser_grad_norm": 0.8695882111787796, "chaser_new_fields": 46.95000076293945, "chaser_partner_vision": 0.7120000123977661, "chaser_return": 11.195000648498535, "chaser_subspace_norm": 35.847537994384766, "collisions_per_episode": 11.050000190734863, "entropy": 0.40138236433267593, "explorer_grad_norm": 2.2347845435142517, "explorer_new_fields": 62.20000076293945, "explorer_partner_vision": 0.7120000123977661, "explorer_return": 34.212501525878906, "explorer_subspace_norm": 19.718151092529297, "final_distance": 2.626574754714966, "max_param_abs": 1.5664823055267334, "policy_loss": -0.006639687702062247, "value_loss": 3.6984063386917114}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:08.736773+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024688.7367861, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8035011398275189, "elapsed_seconds": 2892.604103379068, "env_steps_per_second": 4148.511020219427, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.553173125689273, "eta_seconds": 16391.423252481385, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004280191897279906, "chaser_grad_norm": 1.6289285719394684, "chaser_new_fields": 46.525001525878906, "chaser_partner_vision": 0.7922500371932983, "chaser_return": 14.697501182556152, "chaser_subspace_norm": 42.03453063964844, "collisions_per_episode": 14.574999809265137, "entropy": 0.36487194895744324, "explorer_grad_norm": 1.4224601089954376, "explorer_new_fields": 62.400001525878906, "explorer_partner_vision": 0.7922500371932983, "explorer_return": 31.4375, "explorer_subspace_norm": 26.108469009399414, "final_distance": 2.3828465938568115, "max_param_abs": 1.6350244283676147, "policy_loss": -0.007981847560074584, "value_loss": 3.959980845451355}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:46:09.298016+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025169.2980378, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.940343225768916, "elapsed_seconds": 3385.235612768098, "env_steps_per_second": 4135.605789799735, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.433046635767747, "eta_seconds": 15958.96788876389, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.003137840197549508, "chaser_grad_norm": 1.5023635923862457, "chaser_new_fields": 48.25, "chaser_partner_vision": 0.8252500295639038, "chaser_return": 13.160000801086426, "chaser_subspace_norm": 44.00979995727539, "collisions_per_episode": 12.800000190734863, "entropy": 0.37482111901044846, "explorer_grad_norm": 1.562868446111679, "explorer_new_fields": 60.70000076293945, "explorer_partner_vision": 0.8252500295639038, "explorer_return": 31.087499618530273, "explorer_subspace_norm": 28.60947036743164, "final_distance": 2.096140146255493, "max_param_abs": 1.695357322692871, "policy_loss": -0.006578959751877278, "value_loss": 3.912420392036438}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:54:21.929524+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025661.929536, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0759297312139016, "elapsed_seconds": 3873.347032370046, "env_steps_per_second": 4130.794340472465, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.303718924855606, "eta_seconds": 15493.388129480181, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0026318452445259766, "chaser_grad_norm": 1.6901334524154663, "chaser_new_fields": 50.775001525878906, "chaser_partner_vision": 0.9317500591278076, "chaser_return": 21.380001068115234, "chaser_subspace_norm": 55.47392272949219, "collisions_per_episode": 19.950000762939453, "entropy": 0.33022114634513855, "explorer_grad_norm": 1.2606963217258453, "explorer_new_fields": 62.60000228881836, "explorer_partner_vision": 0.9317500591278076, "explorer_return": 27.0625, "explorer_subspace_norm": 29.616779327392578, "final_distance": 2.113325595855713, "max_param_abs": 1.7328490018844604, "policy_loss": -0.0074894526023285835, "value_loss": 4.052980065345764}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:02:30.040946+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026150.0409613, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.213011313378617, "elapsed_seconds": 4366.840728163021, "env_steps_per_second": 4121.973096914844, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.178150079415236, "eta_seconds": 15041.34028589485, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0036019308668899797, "chaser_grad_norm": 1.371612012386322, "chaser_new_fields": 49.10000228881836, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 22.47749900817871, "chaser_subspace_norm": 67.29763793945312, "collisions_per_episode": 21.17500114440918, "entropy": 0.29616526514291763, "explorer_grad_norm": 1.5929699838161469, "explorer_new_fields": 57.150001525878906, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 18.950000762939453, "explorer_subspace_norm": 32.37251663208008, "final_distance": 1.7971737384796143, "max_param_abs": 1.7635774612426758, "policy_loss": -0.0062066453276230504, "value_loss": 4.3600897789001465}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:10:43.534639+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026643.534652, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.34919901123196, "elapsed_seconds": 4857.116440435057, "env_steps_per_second": 4117.669453732219, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.04759703369588, "eta_seconds": 14571.34932130517, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.002354301818656701, "chaser_grad_norm": 1.88572496175766, "chaser_new_fields": 47.67499923706055, "chaser_partner_vision": 0.9515000581741333, "chaser_return": 19.25750160217285, "chaser_subspace_norm": 76.89038848876953, "collisions_per_episode": 18.42500114440918, "entropy": 0.29250165075063705, "explorer_grad_norm": 2.3060189485549927, "explorer_new_fields": 56.025001525878906, "explorer_partner_vision": 0.9515000581741333, "explorer_return": 19.200000762939453, "explorer_subspace_norm": 33.92460250854492, "final_distance": 1.7689491510391235, "max_param_abs": 1.7871687412261963, "policy_loss": -0.0028539749102236556, "value_loss": 4.3047873973846436}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:18:53.810354+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027133.8103685, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4826748294389107, "elapsed_seconds": 5337.629385980079, "env_steps_per_second": 4121.679946117209, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.908870004884401, "eta_seconds": 14071.932017583844, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.004147187740452196, "chaser_grad_norm": 0.7124435752630234, "chaser_new_fields": 50.775001525878906, "chaser_partner_vision": 0.9500000476837158, "chaser_return": 18.082502365112305, "chaser_subspace_norm": 74.88532257080078, "collisions_per_episode": 16.92500114440918, "entropy": 0.26584532111883163, "explorer_grad_norm": 1.3961082696914673, "explorer_new_fields": 61.025001525878906, "explorer_partner_vision": 0.9500000476837158, "explorer_return": 27.30000114440918, "explorer_subspace_norm": 37.285240173339844, "final_distance": 1.8500149250030518, "max_param_abs": 1.8247284889221191, "policy_loss": -0.006241507373747623, "value_loss": 3.837724447250366}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:26:54.323299+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027614.3233101, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.617673119030563, "elapsed_seconds": 5823.623228510027, "env_steps_per_second": 4121.145729776272, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.7745706110713133, "eta_seconds": 13588.454199856727, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.004896613093276159, "chaser_grad_norm": 0.5519395917654037, "chaser_new_fields": 52.47500228881836, "chaser_partner_vision": 0.971250057220459, "chaser_return": 18.01500129699707, "chaser_subspace_norm": 86.55854034423828, "collisions_per_episode": 16.55000114440918, "entropy": 0.2513393759727478, "explorer_grad_norm": 1.1996820718050003, "explorer_new_fields": 59.375, "explorer_partner_vision": 0.971250057220459, "explorer_return": 25.3125, "explorer_subspace_norm": 35.70689392089844, "final_distance": 1.6991291046142578, "max_param_abs": 1.8441742658615112, "policy_loss": -0.00678599316159989, "value_loss": 3.6564791798591614}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:00.317142+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028100.3171573, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7536866522122485, "elapsed_seconds": 6313.271947964095, "env_steps_per_second": 4118.3082582692305, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.642272277671593, "eta_seconds": 13112.180199617735, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.004086122180028795, "chaser_grad_norm": 0.75764000415802, "chaser_new_fields": 52.775001525878906, "chaser_partner_vision": 0.9587500691413879, "chaser_return": 15.160000801086426, "chaser_subspace_norm": 92.97501373291016, "collisions_per_episode": 13.850000381469727, "entropy": 0.23363753408193588, "explorer_grad_norm": 1.593760997056961, "explorer_new_fields": 61.45000076293945, "explorer_partner_vision": 0.9587500691413879, "explorer_return": 29.55000114440918, "explorer_subspace_norm": 37.97260665893555, "final_distance": 1.4861637353897095, "max_param_abs": 1.8683383464813232, "policy_loss": -0.006644969807922507, "value_loss": 3.4321812987327576}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:43:09.965859+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028589.9658716, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8895175237569493, "elapsed_seconds": 6802.263085525017, "env_steps_per_second": 4116.277134235375, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5091039726914772, "eta_seconds": 12632.774301689318, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004033661646076325, "chaser_grad_norm": 0.8511547893285751, "chaser_new_fields": 50.57500076293945, "chaser_partner_vision": 0.9457500576972961, "chaser_return": 15.735000610351562, "chaser_subspace_norm": 99.34052276611328, "collisions_per_episode": 14.800000190734863, "entropy": 0.26657039672136307, "explorer_grad_norm": 1.5675743222236633, "explorer_new_fields": 61.29999923706055, "explorer_partner_vision": 0.9457500576972961, "explorer_return": 29.712499618530273, "explorer_subspace_norm": 36.880069732666016, "final_distance": 1.7168331146240234, "max_param_abs": 1.9060200452804565, "policy_loss": -0.006028186019509718, "value_loss": 3.638098359107971}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:51:18.956999+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029078.9570138, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0237335863019688, "elapsed_seconds": 7285.440910687088, "env_steps_per_second": 4117.8015672315305, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3728893105032816, "eta_seconds": 12142.401517811813, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004756460233796389, "chaser_grad_norm": 0.5709119439125061, "chaser_new_fields": 52.375, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 13.895001411437988, "chaser_subspace_norm": 110.31792449951172, "collisions_per_episode": 12.75, "entropy": 0.23116421326994896, "explorer_grad_norm": 1.9448820352554321, "explorer_new_fields": 61.54999923706055, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 32.23749923706055, "explorer_subspace_norm": 37.94101333618164, "final_distance": 1.506874442100525, "max_param_abs": 1.9142425060272217, "policy_loss": -0.0029434214783634616, "value_loss": 3.6089770197868347}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:59:22.134822+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029562.134835, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.157074145655303, "elapsed_seconds": 7765.466924359091, "env_steps_per_second": 4120.808228494395, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2356112184829544, "eta_seconds": 11648.200386538636, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004825972480148266, "chaser_grad_norm": 0.8283534049987793, "chaser_new_fields": 53.57500076293945, "chaser_partner_vision": 0.968000054359436, "chaser_return": 14.785000801086426, "chaser_subspace_norm": 120.99627685546875, "collisions_per_episode": 13.40000057220459, "entropy": 0.21304067969322205, "explorer_grad_norm": 1.6758573651313782, "explorer_new_fields": 60.97500228881836, "explorer_partner_vision": 0.968000054359436, "explorer_return": 30.225000381469727, "explorer_subspace_norm": 42.882568359375, "final_distance": 1.833226203918457, "max_param_abs": 1.9629096984863281, "policy_loss": -0.004701275992673004, "value_loss": 3.5094359517097473}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:07:22.160837+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030042.1608524, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.29144471667002, "elapsed_seconds": 8249.200980012072, "env_steps_per_second": 4121.611303007706, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1001899107888504, "eta_seconds": 11160.683678839861, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004641206112097507, "chaser_grad_norm": 0.6338439509272575, "chaser_new_fields": 51.900001525878906, "chaser_partner_vision": 0.9535000324249268, "chaser_return": 16.375, "chaser_subspace_norm": 113.87093353271484, "collisions_per_episode": 15.25, "entropy": 0.18218044564127922, "explorer_grad_norm": 1.5297481417655945, "explorer_new_fields": 59.45000076293945, "explorer_partner_vision": 0.9535000324249268, "explorer_return": 27.05000114440918, "explorer_subspace_norm": 51.30567169189453, "final_distance": 1.5738579034805298, "max_param_abs": 1.9855176210403442, "policy_loss": -0.003959622513519889, "value_loss": 3.634140968322754}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:15:25.894892+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030525.8949041, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.429534612166398, "elapsed_seconds": 8746.324603799032, "env_steps_per_second": 4116.014626802569, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.969431192647819, "eta_seconds": 10689.95229353215, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0026175260243093135, "chaser_grad_norm": 0.6224627643823624, "chaser_new_fields": 51.57500076293945, "chaser_partner_vision": 0.9632500410079956, "chaser_return": 15.824999809265137, "chaser_subspace_norm": 104.75340270996094, "collisions_per_episode": 14.600000381469727, "entropy": 0.25375792384147644, "explorer_grad_norm": 1.8279874622821808, "explorer_new_fields": 58.35000228881836, "explorer_partner_vision": 0.9632500410079956, "explorer_return": 26.4375, "explorer_subspace_norm": 45.77794647216797, "final_distance": 1.6816809177398682, "max_param_abs": 2.0315608978271484, "policy_loss": -0.004086149799633176, "value_loss": 4.016404688358307}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:23:43.018517+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031023.0185318, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5654399140622184, "elapsed_seconds": 9235.583690623986, "env_steps_per_second": 4114.520670586073, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8354862208056097, "eta_seconds": 10207.750394900195, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.004043229913665375, "chaser_grad_norm": 0.8833614736795425, "chaser_new_fields": 57.57500076293945, "chaser_partner_vision": 0.9615000486373901, "chaser_return": 11.220000267028809, "chaser_subspace_norm": 128.8538818359375, "collisions_per_episode": 9.40000057220459, "entropy": 0.1913204789161682, "explorer_grad_norm": 1.1261723190546036, "explorer_new_fields": 63.32500076293945, "explorer_partner_vision": 0.9615000486373901, "explorer_return": 36.61249923706055, "explorer_subspace_norm": 50.340911865234375, "final_distance": 1.6229671239852905, "max_param_abs": 2.0584535598754883, "policy_loss": -0.004367929488424238, "value_loss": 2.852793276309967}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:31:52.277604+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031512.2776184, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.702483517024472, "elapsed_seconds": 9728.9406612881, "env_steps_per_second": 4111.444543922632, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.702483517024472, "eta_seconds": 9728.9406612881, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.003193463740508662, "chaser_grad_norm": 0.3281538635492325, "chaser_new_fields": 55.07500076293945, "chaser_partner_vision": 0.9510000348091125, "chaser_return": 14.852500915527344, "chaser_subspace_norm": 127.62368774414062, "collisions_per_episode": 13.225000381469727, "entropy": 0.1898086778819561, "explorer_grad_norm": 1.373610645532608, "explorer_new_fields": 62.29999923706055, "explorer_partner_vision": 0.9510000348091125, "explorer_return": 32.60000228881836, "explorer_subspace_norm": 59.41374969482422, "final_distance": 1.754033088684082, "max_param_abs": 2.078979730606079, "policy_loss": -0.0022539431043924196, "value_loss": 3.4202033281326294}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:40:05.634576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032005.634592, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.836176120979168, "elapsed_seconds": 10210.234035525005, "env_steps_per_second": 4113.519812951122, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.566064109457343, "eta_seconds": 9237.830794046435, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.003856530135839975, "chaser_grad_norm": 0.3630855157971382, "chaser_new_fields": 49.10000228881836, "chaser_partner_vision": 0.9600000381469727, "chaser_return": 28.0625, "chaser_subspace_norm": 152.66452026367188, "collisions_per_episode": 26.17500114440918, "entropy": 0.18451018258929253, "explorer_grad_norm": 1.062723457813263, "explorer_new_fields": 52.5, "explorer_partner_vision": 0.9600000381469727, "explorer_return": 13.0, "explorer_subspace_norm": 60.721435546875, "final_distance": 1.4646209478378296, "max_param_abs": 2.1176655292510986, "policy_loss": -0.002414814134127141, "value_loss": 4.72590708732605}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:48:06.927948+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032486.9279594, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9703352949397392, "elapsed_seconds": 10693.207061783061, "env_steps_per_second": 4114.761805862116, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.430274332223423, "eta_seconds": 8748.987596004323, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0033079387786010273, "chaser_grad_norm": 0.7346276789903641, "chaser_new_fields": 57.275001525878906, "chaser_partner_vision": 0.9512500166893005, "chaser_return": 13.985000610351562, "chaser_subspace_norm": 162.62796020507812, "collisions_per_episode": 12.050000190734863, "entropy": 0.18784821033477783, "explorer_grad_norm": 1.1042179316282272, "explorer_new_fields": 64.42500305175781, "explorer_partner_vision": 0.9512500166893005, "explorer_return": 36.337501525878906, "explorer_subspace_norm": 54.16365051269531, "final_distance": 1.7238304615020752, "max_param_abs": 2.121508836746216, "policy_loss": -0.00574195051751758, "value_loss": 3.0925402641296387}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:56:09.900975+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032969.9009879, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.103811815577242, "elapsed_seconds": 11173.722536078072, "env_steps_per_second": 4116.801706098727, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.294121776731005, "eta_seconds": 8258.838396231618, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0023447146629136384, "chaser_grad_norm": 0.7196819484233856, "chaser_new_fields": 58.17499923706055, "chaser_partner_vision": 0.9547500610351562, "chaser_return": 11.912500381469727, "chaser_subspace_norm": 162.3022003173828, "collisions_per_episode": 9.875, "entropy": 0.19151673838496208, "explorer_grad_norm": 1.143787682056427, "explorer_new_fields": 66.1500015258789, "explorer_partner_vision": 0.9547500610351562, "explorer_return": 40.275001525878906, "explorer_subspace_norm": 56.355350494384766, "final_distance": 1.5999761819839478, "max_param_abs": 2.2224984169006348, "policy_loss": -0.005386231520219553, "value_loss": 3.174909234046936}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:04:10.416448+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033450.416461, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2370662833672315, "elapsed_seconds": 11653.438620122033, "env_steps_per_second": 4118.9559206257145, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1580441889114876, "eta_seconds": 7768.959080081355, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0032287922477891584, "chaser_grad_norm": 0.40384627878665924, "chaser_new_fields": 56.35000228881836, "chaser_partner_vision": 0.9570000171661377, "chaser_return": 12.282500267028809, "chaser_subspace_norm": 165.6447296142578, "collisions_per_episode": 10.625, "entropy": 0.1941794864833355, "explorer_grad_norm": 1.3268453180789948, "explorer_new_fields": 64.5250015258789, "explorer_partner_vision": 0.9570000171661377, "explorer_return": 37.76250076293945, "explorer_subspace_norm": 54.868934631347656, "final_distance": 1.4310423135757446, "max_param_abs": 2.323395013809204, "policy_loss": -0.005261918940590249, "value_loss": 2.9778467416763306}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:12:10.132533+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033930.1325476, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3719771295205767, "elapsed_seconds": 12139.117666274076, "env_steps_per_second": 4118.915507254224, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.023186277712346, "eta_seconds": 7283.470599764445, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.007444660168342665, "chaser_grad_norm": 0.5806861594319344, "chaser_new_fields": 55.70000076293945, "chaser_partner_vision": 0.9545000195503235, "chaser_return": 14.08750057220459, "chaser_subspace_norm": 146.84739685058594, "collisions_per_episode": 12.475000381469727, "entropy": 0.1622682437300682, "explorer_grad_norm": 0.9118330478668213, "explorer_new_fields": 62.025001525878906, "explorer_partner_vision": 0.9545000195503235, "explorer_return": 31.962499618530273, "explorer_subspace_norm": 66.09556579589844, "final_distance": 1.6498632431030273, "max_param_abs": 2.4275622367858887, "policy_loss": -0.0050485248128069315, "value_loss": 3.114036798477173}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:20:15.811578+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034415.81159, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.511711760631135, "elapsed_seconds": 12642.162338272086, "env_steps_per_second": 4113.220397635496, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.890921717262919, "eta_seconds": 6807.318182146508, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0027336882797817097, "chaser_grad_norm": 0.6803714334964752, "chaser_new_fields": 54.07500076293945, "chaser_partner_vision": 0.968500018119812, "chaser_return": 13.610000610351562, "chaser_subspace_norm": 167.95628356933594, "collisions_per_episode": 12.15000057220459, "entropy": 0.17988111451268196, "explorer_grad_norm": 1.0405263453722, "explorer_new_fields": 62.20000076293945, "explorer_partner_vision": 0.968500018119812, "explorer_return": 33.32500076293945, "explorer_subspace_norm": 69.7338638305664, "final_distance": 1.564717173576355, "max_param_abs": 2.493201732635498, "policy_loss": -0.004828489489852927, "value_loss": 3.362410843372345}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:28:38.856254+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034918.8562686, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6455998685041817, "elapsed_seconds": 13124.159526615054, "env_steps_per_second": 4114.549193835312, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7552888255760877, "eta_seconds": 6319.039772073916, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.004788360669770952, "chaser_grad_norm": 0.41351383179426193, "chaser_new_fields": 54.45000076293945, "chaser_partner_vision": 0.9737500548362732, "chaser_return": 12.449999809265137, "chaser_subspace_norm": 147.37478637695312, "collisions_per_episode": 11.15000057220459, "entropy": 0.20082449913024902, "explorer_grad_norm": 1.2622922956943512, "explorer_new_fields": 63.17499923706055, "explorer_partner_vision": 0.9737500548362732, "explorer_return": 35.66250228881836, "explorer_subspace_norm": 66.22734832763672, "final_distance": 1.6235283613204956, "max_param_abs": 2.6679258346557617, "policy_loss": -0.005181499708456272, "value_loss": 3.0976051092147827}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:36:40.853438+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035400.8534505, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7798569175572223, "elapsed_seconds": 13607.484903206001, "env_steps_per_second": 4115.382114942202, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6199386789530954, "eta_seconds": 5831.779244231144, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0037923446912584424, "chaser_grad_norm": 0.7345854938030243, "chaser_new_fields": 57.97500228881836, "chaser_partner_vision": 0.9767500758171082, "chaser_return": 13.830000877380371, "chaser_subspace_norm": 174.7973175048828, "collisions_per_episode": 11.850000381469727, "entropy": 0.1755659021437168, "explorer_grad_norm": 1.0903593301773071, "explorer_new_fields": 65.05000305175781, "explorer_partner_vision": 0.9767500758171082, "explorer_return": 38.04999923706055, "explorer_subspace_norm": 68.35354614257812, "final_distance": 1.7706577777862549, "max_param_abs": 2.693056344985962, "policy_loss": -0.004521974256306294, "value_loss": 3.107781410217285}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:44:44.178818+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035884.178832, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9167887260769687, "elapsed_seconds": 14100.439413877088, "env_steps_per_second": 4113.3469885285085, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4856784823050568, "eta_seconds": 5348.442536298205, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0031074752979094455, "chaser_grad_norm": 0.4263124465942383, "chaser_new_fields": 56.25, "chaser_partner_vision": 0.967750072479248, "chaser_return": 16.287500381469727, "chaser_subspace_norm": 176.8242950439453, "collisions_per_episode": 14.175000190734863, "entropy": 0.20270903781056404, "explorer_grad_norm": 1.110108956694603, "explorer_new_fields": 63.82500076293945, "explorer_partner_vision": 0.967750072479248, "explorer_return": 34.150001525878906, "explorer_subspace_norm": 66.77369689941406, "final_distance": 1.5881727933883667, "max_param_abs": 2.728546380996704, "policy_loss": -0.005597370059263085, "value_loss": 3.456786811351776}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:52:57.133327+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036377.1333406, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.055110073168083, "elapsed_seconds": 14598.396263405099, "env_steps_per_second": 4110.040508381495, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.351703357722694, "eta_seconds": 4866.132087801699, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0021531521779145635, "chaser_grad_norm": 0.6278064996004105, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.9667500257492065, "chaser_return": 13.100000381469727, "chaser_subspace_norm": 166.13958740234375, "collisions_per_episode": 11.949999809265137, "entropy": 0.21180688589811325, "explorer_grad_norm": 0.9927292466163635, "explorer_new_fields": 62.5, "explorer_partner_vision": 0.9667500257492065, "explorer_return": 34.13750076293945, "explorer_subspace_norm": 75.1683578491211, "final_distance": 1.7419010400772095, "max_param_abs": 2.808729410171509, "policy_loss": -0.0042310497472186626, "value_loss": 3.6787883043289185}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:01:15.090178+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036875.0901942, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.190319242341664, "elapsed_seconds": 15085.14927242999, "env_steps_per_second": 4110.002418956026, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.216544296163709, "eta_seconds": 4379.559466189352, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0030292084457843593, "chaser_grad_norm": 0.5486030727624893, "chaser_new_fields": 58.70000076293945, "chaser_partner_vision": 0.9702500700950623, "chaser_return": 12.449999809265137, "chaser_subspace_norm": 177.70956420898438, "collisions_per_episode": 10.350000381469727, "entropy": 0.1710856705904007, "explorer_grad_norm": 1.0047283917665482, "explorer_new_fields": 64.95000457763672, "explorer_partner_vision": 0.9702500700950623, "explorer_return": 38.20000076293945, "explorer_subspace_norm": 75.10011291503906, "final_distance": 1.5267529487609863, "max_param_abs": 2.827589988708496, "policy_loss": -0.0055586954280784084, "value_loss": 2.6424641013145447}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:09:21.843184+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037361.8431952, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.324301193853349, "elapsed_seconds": 15567.484297872055, "env_steps_per_second": 4111.133101238988, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0810752984633372, "eta_seconds": 3891.871074468014, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.004711658318566236, "chaser_grad_norm": 0.6172671318054199, "chaser_new_fields": 54.95000076293945, "chaser_partner_vision": 0.968250036239624, "chaser_return": 15.405000686645508, "chaser_subspace_norm": 191.79356384277344, "collisions_per_episode": 13.75, "entropy": 0.20827138051390648, "explorer_grad_norm": 1.2041530013084412, "explorer_new_fields": 63.150001525878906, "explorer_partner_vision": 0.968250036239624, "explorer_return": 33.837501525878906, "explorer_subspace_norm": 75.63551330566406, "final_distance": 1.6731315851211548, "max_param_abs": 2.8288967609405518, "policy_loss": -0.003963908256949367, "value_loss": 3.4593767523765564}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:17:24.178210+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037844.1782248, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.458616231236956, "elapsed_seconds": 16051.018432453042, "env_steps_per_second": 4111.888618017952, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9457670793532936, "eta_seconds": 3404.761485671857, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0029192062141119024, "chaser_grad_norm": 0.9972567558288574, "chaser_new_fields": 54.75, "chaser_partner_vision": 0.971750020980835, "chaser_return": 13.167498588562012, "chaser_subspace_norm": 191.86756896972656, "collisions_per_episode": 11.574999809265137, "entropy": 0.17266111075878143, "explorer_grad_norm": 1.241422861814499, "explorer_new_fields": 63.95000076293945, "explorer_partner_vision": 0.971750020980835, "explorer_return": 36.5, "explorer_subspace_norm": 75.6630630493164, "final_distance": 1.6068744659423828, "max_param_abs": 2.813716173171997, "policy_loss": -0.003948502249087937, "value_loss": 3.558075785636902}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:25:27.712344+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038327.7123563, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.593835567008064, "elapsed_seconds": 16537.80804122903, "env_steps_per_second": 4111.790379382496, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8106768647661289, "eta_seconds": 2918.436713158064, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.002837662649948225, "chaser_grad_norm": 0.5180001333355904, "chaser_new_fields": 55.625, "chaser_partner_vision": 0.9572500586509705, "chaser_return": 13.297500610351562, "chaser_subspace_norm": 203.66506958007812, "collisions_per_episode": 11.625, "entropy": 0.1751151643693447, "explorer_grad_norm": 1.2901441603899002, "explorer_new_fields": 62.04999923706055, "explorer_partner_vision": 0.9572500586509705, "explorer_return": 33.92499923706055, "explorer_subspace_norm": 84.80243682861328, "final_distance": 1.4088834524154663, "max_param_abs": 2.863408327102661, "policy_loss": -0.005207648817460608, "value_loss": 3.3028537034988403}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:33:34.501954+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038814.501968, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.730334096042239, "elapsed_seconds": 17029.20274575206, "env_steps_per_second": 4110.585859191883, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6757620137203197, "eta_seconds": 2432.743249393151, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0031371789733256605, "chaser_grad_norm": 0.6977758407592773, "chaser_new_fields": 54.29999923706055, "chaser_partner_vision": 0.9565000534057617, "chaser_return": 14.792502403259277, "chaser_subspace_norm": 243.31515502929688, "collisions_per_episode": 13.324999809265137, "entropy": 0.18260349333286285, "explorer_grad_norm": 0.99324531853199, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9565000534057617, "explorer_return": 34.275001525878906, "explorer_subspace_norm": 76.68460845947266, "final_distance": 1.4965190887451172, "max_param_abs": 2.821523427963257, "policy_loss": -0.004713630930510249, "value_loss": 3.5730031728744507}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:41:45.896657+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039305.8966699, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.864364715716963, "elapsed_seconds": 17511.712976581068, "env_steps_per_second": 4111.533811471655, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5404849684129959, "eta_seconds": 1945.7458862867852, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.003807815049164276, "chaser_grad_norm": 0.9422713667154312, "chaser_new_fields": 56.75, "chaser_partner_vision": 0.9810000658035278, "chaser_return": 10.875, "chaser_subspace_norm": 200.8000030517578, "collisions_per_episode": 9.100000381469727, "entropy": 0.19070982560515404, "explorer_grad_norm": 0.9053702354431152, "explorer_new_fields": 63.5, "explorer_partner_vision": 0.9810000658035278, "explorer_return": 37.212501525878906, "explorer_subspace_norm": 79.01789093017578, "final_distance": 1.5713547468185425, "max_param_abs": 2.840312957763672, "policy_loss": -0.004082289715313436, "value_loss": 3.0113396048545837}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:49:48.406889+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039788.4069033, "update": 18000} +{"device": "cuda", "elapsed_hours": 4.99901422015139, "elapsed_seconds": 17996.451192545006, "env_steps_per_second": 4111.921801041216, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4053254773095722, "eta_seconds": 1459.17171831446, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0032483166336447766, "chaser_grad_norm": 0.38564686477184296, "chaser_new_fields": 59.17499923706055, "chaser_partner_vision": 0.9620000720024109, "chaser_return": 15.3774995803833, "chaser_subspace_norm": 242.44549560546875, "collisions_per_episode": 12.725000381469727, "entropy": 0.1465696468949318, "explorer_grad_norm": 0.6050713658332825, "explorer_new_fields": 67.25, "explorer_partner_vision": 0.9620000720024109, "explorer_return": 40.35000228881836, "explorer_subspace_norm": 80.4219741821289, "final_distance": 1.686489462852478, "max_param_abs": 2.846783399581909, "policy_loss": -0.00575739636644812, "value_loss": 2.7126697301864624}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:57:53.145104+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040273.1451185, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.137936197075017, "elapsed_seconds": 18496.57030947006, "env_steps_per_second": 4108.869846054042, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2704176945828956, "eta_seconds": 973.5037004984241, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.001569945413467444, "chaser_grad_norm": 0.6780674308538437, "chaser_new_fields": 58.900001525878906, "chaser_partner_vision": 0.9672500491142273, "chaser_return": 11.572501182556152, "chaser_subspace_norm": 237.07382202148438, "collisions_per_episode": 9.475000381469727, "entropy": 0.14992974698543549, "explorer_grad_norm": 1.299996167421341, "explorer_new_fields": 66.9000015258789, "explorer_partner_vision": 0.9672500491142273, "explorer_return": 42.087501525878906, "explorer_subspace_norm": 87.1292724609375, "final_distance": 1.5733639001846313, "max_param_abs": 2.847513437271118, "policy_loss": -0.0035836435556433344, "value_loss": 2.838916301727295}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:06:13.264225+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040773.2642395, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.273635968858337, "elapsed_seconds": 18985.089487890014, "env_steps_per_second": 4108.487350020327, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13522143509893172, "eta_seconds": 486.7971663561542, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.005018933984096918, "chaser_grad_norm": 0.7915496677160263, "chaser_new_fields": 58.07500076293945, "chaser_partner_vision": 0.9667500257492065, "chaser_return": 11.980000495910645, "chaser_subspace_norm": 244.86062622070312, "collisions_per_episode": 9.850000381469727, "entropy": 0.13851149380207062, "explorer_grad_norm": 0.8033619523048401, "explorer_new_fields": 64.375, "explorer_partner_vision": 0.9667500257492065, "explorer_return": 38.26250076293945, "explorer_subspace_norm": 97.4024429321289, "final_distance": 1.6643885374069214, "max_param_abs": 2.818692445755005, "policy_loss": -0.0024935723040329094, "value_loss": 2.5776590704917908}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:14:21.783401+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041261.7834163, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.411322240413897, "elapsed_seconds": 19480.76006549003, "env_steps_per_second": 4106.615949842696, "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-transformer/social/seed_0006/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:22:37.453977+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041757.4539888, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.411338810608334, "elapsed_seconds": 19480.819718190003, "env_steps_per_second": 4106.603374872407, "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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0015035698563394817, "chaser_grad_norm": 0.5757317766547203, "chaser_new_fields": 63.22500228881836, "chaser_partner_vision": 0.9730000495910645, "chaser_return": 12.269999504089355, "chaser_subspace_norm": 190.25613403320312, "collisions_per_episode": 9.300000190734863, "entropy": 0.18287497758865356, "explorer_grad_norm": 0.6242441534996033, "explorer_new_fields": 69.2249984741211, "explorer_partner_vision": 0.9730000495910645, "explorer_return": 46.0, "explorer_subspace_norm": 97.26615905761719, "final_distance": 1.7034721374511719, "max_param_abs": 2.8255860805511475, "policy_loss": -0.005794699959466243, "value_loss": 2.655023217201233}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:22:37.513633+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041757.5136476, "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..68bf0ae96a03c7fb8bc033bd755267c2b1337dd4 --- /dev/null +++ b/social/seed_0006/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.411338810608334, + "elapsed_seconds": 19480.819718190003, + "env_steps_per_second": 4106.603374872407, + "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-transformer/social/seed_0006/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0015035698563394817, + "chaser_grad_norm": 0.5757317766547203, + "chaser_new_fields": 63.22500228881836, + "chaser_partner_vision": 0.9730000495910645, + "chaser_return": 12.269999504089355, + "chaser_subspace_norm": 190.25613403320312, + "collisions_per_episode": 9.300000190734863, + "entropy": 0.18287497758865356, + "explorer_grad_norm": 0.6242441534996033, + "explorer_new_fields": 69.2249984741211, + "explorer_partner_vision": 0.9730000495910645, + "explorer_return": 46.0, + "explorer_subspace_norm": 97.26615905761719, + "final_distance": 1.7034721374511719, + "max_param_abs": 2.8255860805511475, + "policy_loss": -0.005794699959466243, + "value_loss": 2.655023217201233 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:22:37.513633+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041757.5136476, + "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..2c9425e3595c63a677342bc45b17e053df83de45 --- /dev/null +++ b/social/seed_0007/attempt_01/checkpoints/latest.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:31976e8959949de1246f21e9902b4882a9b47763a935458165086ad52ff66214 +size 29760376 diff --git a/social/seed_0007/attempt_01/config.json b/social/seed_0007/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..cb52e205869ab7acdfff841561d60bf6eeaaf4f8 --- /dev/null +++ b/social/seed_0007/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0007/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..06f5a91d3cf341fd90c21dce97843a60f95d0ae4 --- /dev/null +++ b/social/seed_0007/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0006420283175086499, "elapsed_seconds": 2.3113019430311397, "env_steps_per_second": 1730.6263303505166, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 12.83992432185549, "eta_seconds": 46223.72755867976, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0042613881821580435, "chaser_grad_norm": 4.248058281838894, "chaser_new_fields": 33.92499923706055, "chaser_partner_vision": 0.2997500002384186, "chaser_return": -1.4350003004074097, "chaser_subspace_norm": 0.03431187942624092, "collisions_per_episode": 1.649999976158142, "entropy": 2.7583693265914917, "explorer_grad_norm": 5.865336924791336, "explorer_new_fields": 34.67499923706055, "explorer_partner_vision": 0.2997500002384186, "explorer_return": 0.699999988079071, "explorer_subspace_norm": 0.03817305341362953, "final_distance": 5.4165120124816895, "max_param_abs": 1.0012027025222778, "policy_loss": -0.013772157102835081, "value_loss": 8.035563707351685}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:10.449498+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022290.4495118, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13450471608805756, "elapsed_seconds": 484.21697791700717, "env_steps_per_second": 4130.379749598106, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.2456839274342455, "eta_seconds": 18884.462138763283, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004535545811567232, "chaser_grad_norm": 1.368838757276535, "chaser_new_fields": 45.45000076293945, "chaser_partner_vision": 0.8542500138282776, "chaser_return": 18.922500610351562, "chaser_subspace_norm": 10.764016151428223, "collisions_per_episode": 18.575000762939453, "entropy": 0.5755577236413956, "explorer_grad_norm": 1.1227055937051773, "explorer_new_fields": 61.22500228881836, "explorer_partner_vision": 0.8542500138282776, "explorer_return": 26.850000381469727, "explorer_subspace_norm": 7.975663661956787, "final_distance": 2.359119176864624, "max_param_abs": 1.2460999488830566, "policy_loss": -0.008062180086980675, "value_loss": 4.335863947868347}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:06:12.355177+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022772.3551924, "update": 500} +{"device": "cuda", "elapsed_hours": 0.27026285100582753, "elapsed_seconds": 972.9462636209792, "env_steps_per_second": 4111.223969465018, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.1349941691107235, "eta_seconds": 18485.979008798604, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.005107831419990383, "chaser_grad_norm": 6.284733355045319, "chaser_new_fields": 47.04999923706055, "chaser_partner_vision": 0.7640000581741333, "chaser_return": 11.562500953674316, "chaser_subspace_norm": 15.787992477416992, "collisions_per_episode": 11.52500057220459, "entropy": 0.4450680539011955, "explorer_grad_norm": 1.4305541813373566, "explorer_new_fields": 65.0250015258789, "explorer_partner_vision": 0.7640000581741333, "explorer_return": 37.35000228881836, "explorer_subspace_norm": 13.46675968170166, "final_distance": 2.6134190559387207, "max_param_abs": 1.3190828561782837, "policy_loss": -0.00488575137917735, "value_loss": 3.542955756187439}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:14:21.084459+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023261.084474, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4063972269377943, "elapsed_seconds": 1463.0300169760594, "env_steps_per_second": 4101.077852388439, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.01223246556613, "eta_seconds": 18044.036876038066, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.006517593050461734, "chaser_grad_norm": 1.9532191753387451, "chaser_new_fields": 46.125, "chaser_partner_vision": 0.8572500348091125, "chaser_return": 15.797500610351562, "chaser_subspace_norm": 23.099096298217773, "collisions_per_episode": 15.725000381469727, "entropy": 0.37133996188640594, "explorer_grad_norm": 2.6707368195056915, "explorer_new_fields": 62.32500076293945, "explorer_partner_vision": 0.8572500348091125, "explorer_return": 30.1875, "explorer_subspace_norm": 19.13649559020996, "final_distance": 2.5607569217681885, "max_param_abs": 1.4007145166397095, "policy_loss": -0.0046410400941852, "value_loss": 4.031577229499817}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:31.168212+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023751.1682248, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5415157033633377, "elapsed_seconds": 1949.4565321080154, "env_steps_per_second": 4103.707812017394, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.873641330270038, "eta_seconds": 17545.10878897214, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0033110887531009325, "chaser_grad_norm": 1.4637092649936676, "chaser_new_fields": 48.525001525878906, "chaser_partner_vision": 0.9025000333786011, "chaser_return": 19.235000610351562, "chaser_subspace_norm": 28.080604553222656, "collisions_per_episode": 18.5, "entropy": 0.35967475175857544, "explorer_grad_norm": 2.1065503358840942, "explorer_new_fields": 62.82500076293945, "explorer_partner_vision": 0.9025000333786011, "explorer_return": 28.42500114440918, "explorer_subspace_norm": 25.554447174072266, "final_distance": 1.7835975885391235, "max_param_abs": 1.474002480506897, "policy_loss": -0.005086446104570541, "value_loss": 3.900077700614929}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:37.594730+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024237.5947454, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.677057080386407, "elapsed_seconds": 2437.405489391065, "env_steps_per_second": 4102.723179842469, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.739399562704849, "eta_seconds": 17061.838425737456, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.004552271467737512, "chaser_grad_norm": 1.4175283461809158, "chaser_new_fields": 51.07500076293945, "chaser_partner_vision": 0.9570000171661377, "chaser_return": 23.989999771118164, "chaser_subspace_norm": 40.6931037902832, "collisions_per_episode": 22.649999618530273, "entropy": 0.32240520417690277, "explorer_grad_norm": 1.5057802498340607, "explorer_new_fields": 59.400001525878906, "explorer_partner_vision": 0.9570000171661377, "explorer_return": 19.75, "explorer_subspace_norm": 32.50859832763672, "final_distance": 1.7651758193969727, "max_param_abs": 1.4812906980514526, "policy_loss": -0.005155267334558289, "value_loss": 3.711029827594757}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:45.543685+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024725.5436966, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8160446329225023, "elapsed_seconds": 2937.7606785210082, "env_steps_per_second": 4084.7438961710463, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.624252919894179, "eta_seconds": 16647.310511619045, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.004248557231674388, "chaser_grad_norm": 0.7733950316905975, "chaser_new_fields": 44.150001525878906, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 26.457502365112305, "chaser_subspace_norm": 51.827152252197266, "collisions_per_episode": 25.92500114440918, "entropy": 0.3263584226369858, "explorer_grad_norm": 1.9514794647693634, "explorer_new_fields": 52.72500228881836, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 9.074999809265137, "explorer_subspace_norm": 32.537044525146484, "final_distance": 1.6205614805221558, "max_param_abs": 1.5092129707336426, "policy_loss": -0.006498619036747755, "value_loss": 4.350559234619141}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:47:05.898875+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025225.898891, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9523843691824975, "elapsed_seconds": 3428.5837290569907, "env_steps_per_second": 4083.318683849266, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.489812026146059, "eta_seconds": 16163.323294125814, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.01413635168607108, "chaser_grad_norm": 0.9689128398895264, "chaser_new_fields": 45.60000228881836, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 31.08500099182129, "chaser_subspace_norm": 67.076171875, "collisions_per_episode": 29.649999618530273, "entropy": 0.2483539655804634, "explorer_grad_norm": 1.3972825407981873, "explorer_new_fields": 51.75, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 7.175000190734863, "explorer_subspace_norm": 35.22257614135742, "final_distance": 1.6475704908370972, "max_param_abs": 1.5491079092025757, "policy_loss": -0.003923462464470617, "value_loss": 4.988186001777649}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:55:16.721926+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025716.7219398, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.090692515704173, "elapsed_seconds": 3926.4930565350223, "env_steps_per_second": 4074.8830494862455, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.362770062816692, "eta_seconds": 15705.97222614009, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.00413981412675879, "chaser_grad_norm": 0.6604360938072205, "chaser_new_fields": 51.5, "chaser_partner_vision": 0.9547500610351562, "chaser_return": 19.375, "chaser_subspace_norm": 74.0802993774414, "collisions_per_episode": 18.05000114440918, "entropy": 0.28549252450466156, "explorer_grad_norm": 1.3267535269260406, "explorer_new_fields": 58.72500228881836, "explorer_partner_vision": 0.9547500610351562, "explorer_return": 23.774999618530273, "explorer_subspace_norm": 29.7405948638916, "final_distance": 1.583719253540039, "max_param_abs": 1.5555052757263184, "policy_loss": -0.00536348890699756, "value_loss": 3.7818683981895447}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:03:34.631253+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026214.631268, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2266936759697273, "elapsed_seconds": 4416.097233491018, "env_steps_per_second": 4075.997209366384, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.225278217229061, "eta_seconds": 15211.001582024619, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.005726862734739768, "chaser_grad_norm": 0.6680938601493835, "chaser_new_fields": 48.82500076293945, "chaser_partner_vision": 0.9537500739097595, "chaser_return": 20.23750114440918, "chaser_subspace_norm": 78.57432556152344, "collisions_per_episode": 19.274999618530273, "entropy": 0.22769475355744362, "explorer_grad_norm": 1.4325279146432877, "explorer_new_fields": 56.70000076293945, "explorer_partner_vision": 0.9537500739097595, "explorer_return": 19.337499618530273, "explorer_subspace_norm": 34.7354621887207, "final_distance": 1.4936779737472534, "max_param_abs": 1.5950497388839722, "policy_loss": -0.006232022838663198, "value_loss": 3.8379551768302917}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:11:44.235428+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026704.2354386, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.36067434971528, "elapsed_seconds": 4898.427658975008, "env_steps_per_second": 4082.9428119358986, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.08202304914584, "eta_seconds": 14695.282976925024, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.003983062483808197, "chaser_grad_norm": 0.5472320914268494, "chaser_new_fields": 48.20000076293945, "chaser_partner_vision": 0.9590000510215759, "chaser_return": 21.072500228881836, "chaser_subspace_norm": 94.64756774902344, "collisions_per_episode": 20.225000381469727, "entropy": 0.2360350713133812, "explorer_grad_norm": 1.3132319301366806, "explorer_new_fields": 54.97500228881836, "explorer_partner_vision": 0.9590000510215759, "explorer_return": 16.387500762939453, "explorer_subspace_norm": 38.15937042236328, "final_distance": 1.6038007736206055, "max_param_abs": 1.6129662990570068, "policy_loss": -0.005049995197911272, "value_loss": 3.7501031160354614}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:46.565854+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027186.5658703, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4956712767638964, "elapsed_seconds": 5384.416596350027, "env_steps_per_second": 4085.8651269504844, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.943133366013909, "eta_seconds": 14195.280117650072, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.006437666453056323, "chaser_grad_norm": 0.6096645891666412, "chaser_new_fields": 52.32500076293945, "chaser_partner_vision": 0.9580000638961792, "chaser_return": 19.9950008392334, "chaser_subspace_norm": 103.8891372680664, "collisions_per_episode": 18.600000381469727, "entropy": 0.16955404356122017, "explorer_grad_norm": 1.2419367134571075, "explorer_new_fields": 60.35000228881836, "explorer_partner_vision": 0.9580000638961792, "explorer_return": 25.075000762939453, "explorer_subspace_norm": 45.35411071777344, "final_distance": 1.6330902576446533, "max_param_abs": 1.6432785987854004, "policy_loss": -0.004754980430902833, "value_loss": 3.1946789622306824}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:52.554792+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027672.554803, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.630555212810545, "elapsed_seconds": 5869.998766117962, "env_steps_per_second": 4088.586890090958, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.804628829891272, "eta_seconds": 13696.663787608579, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0041217447177057664, "chaser_grad_norm": 0.7462178021669388, "chaser_new_fields": 54.45000076293945, "chaser_partner_vision": 0.9660000205039978, "chaser_return": 13.882501602172852, "chaser_subspace_norm": 101.04923248291016, "collisions_per_episode": 12.375, "entropy": 0.20957104489207268, "explorer_grad_norm": 1.2816796004772186, "explorer_new_fields": 63.32500076293945, "explorer_partner_vision": 0.9660000205039978, "explorer_return": 34.22500228881836, "explorer_subspace_norm": 42.29224395751953, "final_distance": 1.7038007974624634, "max_param_abs": 1.669601559638977, "policy_loss": -0.006279527036483046, "value_loss": 3.4214059114456177}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:58.136961+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028158.1369758, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7666632834955494, "elapsed_seconds": 6359.987820583978, "env_steps_per_second": 4088.0581431070514, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6692237426446024, "eta_seconds": 13209.20547352057, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.003582395577148567, "chaser_grad_norm": 0.89969702064991, "chaser_new_fields": 55.5, "chaser_partner_vision": 0.9652500748634338, "chaser_return": 14.297500610351562, "chaser_subspace_norm": 107.90084838867188, "collisions_per_episode": 12.725000381469727, "entropy": 0.2309948466718197, "explorer_grad_norm": 2.4562712609767914, "explorer_new_fields": 63.75, "explorer_partner_vision": 0.9652500748634338, "explorer_return": 34.462501525878906, "explorer_subspace_norm": 44.59767532348633, "final_distance": 1.642065405845642, "max_param_abs": 1.6976170539855957, "policy_loss": -0.004304959360510452, "value_loss": 3.1022144556045532}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:44:08.126016+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028648.1260293, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.902134899184457, "elapsed_seconds": 6847.685637064045, "env_steps_per_second": 4088.9727543049185, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.532536241342563, "eta_seconds": 12717.130468833227, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004700977416282726, "chaser_grad_norm": 0.4670490548014641, "chaser_new_fields": 52.29999923706055, "chaser_partner_vision": 0.9582500457763672, "chaser_return": 14.287500381469727, "chaser_subspace_norm": 94.9739990234375, "collisions_per_episode": 13.125, "entropy": 0.2529819756746292, "explorer_grad_norm": 1.1683269143104553, "explorer_new_fields": 63.47500228881836, "explorer_partner_vision": 0.9582500457763672, "explorer_return": 34.41250228881836, "explorer_subspace_norm": 51.90651321411133, "final_distance": 1.6971521377563477, "max_param_abs": 1.7332571744918823, "policy_loss": -0.00706003065869254, "value_loss": 3.645749270915985}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:52:15.823833+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029135.823848, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.0406489356119546, "elapsed_seconds": 7346.336168203037, "env_steps_per_second": 4083.6682821361005, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.401081559353258, "eta_seconds": 12243.89361367173, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0043772565645134875, "chaser_grad_norm": 0.4629896208643913, "chaser_new_fields": 56.07500076293945, "chaser_partner_vision": 0.9537500739097595, "chaser_return": 12.360000610351562, "chaser_subspace_norm": 119.24725341796875, "collisions_per_episode": 10.5, "entropy": 0.17080692946910858, "explorer_grad_norm": 1.266626313328743, "explorer_new_fields": 66.2750015258789, "explorer_partner_vision": 0.9537500739097595, "explorer_return": 39.88750076293945, "explorer_subspace_norm": 51.56430435180664, "final_distance": 1.6368759870529175, "max_param_abs": 1.8138337135314941, "policy_loss": -0.005860325882224515, "value_loss": 3.0729271173477173}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:00:34.474363+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029634.474374, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.1769122334730735, "elapsed_seconds": 7836.884040503064, "env_steps_per_second": 4083.2555177052054, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2653683502096102, "eta_seconds": 11755.326060754596, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004455936686202477, "chaser_grad_norm": 0.3577149361371994, "chaser_new_fields": 56.875, "chaser_partner_vision": 0.9640000462532043, "chaser_return": 12.345000267028809, "chaser_subspace_norm": 130.61907958984375, "collisions_per_episode": 10.449999809265137, "entropy": 0.16734926030039787, "explorer_grad_norm": 1.0769377201795578, "explorer_new_fields": 65.7249984741211, "explorer_partner_vision": 0.9640000462532043, "explorer_return": 38.375, "explorer_subspace_norm": 54.723201751708984, "final_distance": 1.5749081373214722, "max_param_abs": 1.859295129776001, "policy_loss": -0.005444348101505625, "value_loss": 2.6681512594223022}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:08:45.022238+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030125.022252, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3133401944255456, "elapsed_seconds": 8328.024699931964, "env_steps_per_second": 4082.600763693432, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.129813204222797, "eta_seconds": 11267.327535202068, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004508690887140787, "chaser_grad_norm": 0.41014116257429123, "chaser_new_fields": 57.0, "chaser_partner_vision": 0.9607500433921814, "chaser_return": 12.8725004196167, "chaser_subspace_norm": 139.48236083984375, "collisions_per_episode": 10.975000381469727, "entropy": 0.17195318639278412, "explorer_grad_norm": 1.2165267616510391, "explorer_new_fields": 67.0250015258789, "explorer_partner_vision": 0.9607500433921814, "explorer_return": 39.98749923706055, "explorer_subspace_norm": 53.66856384277344, "final_distance": 1.6306453943252563, "max_param_abs": 1.8706713914871216, "policy_loss": -0.004337146619796783, "value_loss": 2.7119832038879395}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:16:56.162895+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030616.1629071, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.4501910593761327, "elapsed_seconds": 8820.687813754077, "env_steps_per_second": 4081.314378212693, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9946779614597174, "eta_seconds": 10780.840661254982, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00485014657813565, "chaser_grad_norm": 0.4775408208370209, "chaser_new_fields": 54.22500228881836, "chaser_partner_vision": 0.955750048160553, "chaser_return": 20.030000686645508, "chaser_subspace_norm": 133.76429748535156, "collisions_per_episode": 18.149999618530273, "entropy": 0.200494222342968, "explorer_grad_norm": 1.6115948557853699, "explorer_new_fields": 63.25, "explorer_partner_vision": 0.955750048160553, "explorer_return": 29.762500762939453, "explorer_subspace_norm": 53.62173843383789, "final_distance": 1.5774208307266235, "max_param_abs": 1.9065319299697876, "policy_loss": -0.004449563964271874, "value_loss": 3.641414701938629}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:25:08.826011+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031108.8260443, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.5871209460255664, "elapsed_seconds": 9313.635405692039, "env_steps_per_second": 4080.039463084013, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8594494666598367, "eta_seconds": 10294.018079975413, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0037827110082133686, "chaser_grad_norm": 0.8790079802274704, "chaser_new_fields": 54.22500228881836, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 15.780000686645508, "chaser_subspace_norm": 123.11333465576172, "collisions_per_episode": 14.100000381469727, "entropy": 0.19276471808552742, "explorer_grad_norm": 1.5688157081604004, "explorer_new_fields": 63.92499923706055, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 33.587501525878906, "explorer_subspace_norm": 59.46266555786133, "final_distance": 1.7497825622558594, "max_param_abs": 1.9810999631881714, "policy_loss": -0.0068583547472798045, "value_loss": 3.268198847770691}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:33:21.773601+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031601.7736123, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.721423325336687, "elapsed_seconds": 9797.123971212073, "env_steps_per_second": 4082.8308509248463, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.721423325336687, "eta_seconds": 9797.123971212073, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.003941253607502304, "chaser_grad_norm": 0.6219166070222855, "chaser_new_fields": 54.95000076293945, "chaser_partner_vision": 0.9640000462532043, "chaser_return": 12.510001182556152, "chaser_subspace_norm": 115.0440673828125, "collisions_per_episode": 10.90000057220459, "entropy": 0.22068394720554352, "explorer_grad_norm": 1.477463275194168, "explorer_new_fields": 64.9000015258789, "explorer_partner_vision": 0.9640000462532043, "explorer_return": 37.13750076293945, "explorer_subspace_norm": 65.91960144042969, "final_distance": 1.6271885633468628, "max_param_abs": 2.0652220249176025, "policy_loss": -0.007794993868088795, "value_loss": 3.123426139354706}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:41:25.262166+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032085.2621815, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8577702946030574, "elapsed_seconds": 10287.973060571006, "env_steps_per_second": 4082.4368175074615, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.5856016951170524, "eta_seconds": 9308.166102421388, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0032817334071912896, "chaser_grad_norm": 0.9611087292432785, "chaser_new_fields": 56.525001525878906, "chaser_partner_vision": 0.9565000534057617, "chaser_return": 16.345001220703125, "chaser_subspace_norm": 158.65797424316406, "collisions_per_episode": 14.350000381469727, "entropy": 0.17920908704400063, "explorer_grad_norm": 1.7754017114639282, "explorer_new_fields": 64.95000457763672, "explorer_partner_vision": 0.9565000534057617, "explorer_return": 34.400001525878906, "explorer_subspace_norm": 66.64163208007812, "final_distance": 1.5392242670059204, "max_param_abs": 2.1214165687561035, "policy_loss": -0.005100050441924964, "value_loss": 3.0242255330085754}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:49:36.111257+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032576.1112723, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.993779366231659, "elapsed_seconds": 10777.605718433973, "env_steps_per_second": 4082.5394015613856, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4494558450986306, "eta_seconds": 8818.04104235507, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0033016880011942362, "chaser_grad_norm": 1.2473862767219543, "chaser_new_fields": 46.07500076293945, "chaser_partner_vision": 0.9475000500679016, "chaser_return": 20.705001831054688, "chaser_subspace_norm": 171.10655212402344, "collisions_per_episode": 20.100000381469727, "entropy": 0.2253679521381855, "explorer_grad_norm": 1.06717087328434, "explorer_new_fields": 53.85000228881836, "explorer_partner_vision": 0.9475000500679016, "explorer_return": 15.287500381469727, "explorer_subspace_norm": 67.38903045654297, "final_distance": 1.6106027364730835, "max_param_abs": 2.148735523223877, "policy_loss": -0.004208725180014028, "value_loss": 4.329091787338257}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:45.743915+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033065.7439294, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1299309139869487, "elapsed_seconds": 11267.751290353015, "env_steps_per_second": 4082.447226127835, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.313427197294701, "eta_seconds": 8328.337910260923, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.0019393554085098685, "chaser_grad_norm": 0.5443213880062103, "chaser_new_fields": 54.97500228881836, "chaser_partner_vision": 0.9590000510215759, "chaser_return": 13.315001487731934, "chaser_subspace_norm": 164.77291870117188, "collisions_per_episode": 11.90000057220459, "entropy": 0.19284047186374664, "explorer_grad_norm": 1.0424738377332687, "explorer_new_fields": 61.775001525878906, "explorer_partner_vision": 0.9590000510215759, "explorer_return": 32.625, "explorer_subspace_norm": 63.95738983154297, "final_distance": 1.6821457147598267, "max_param_abs": 2.197927951812744, "policy_loss": -0.004781392747531754, "value_loss": 2.8570324182510376}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:55.889487+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033555.8895013, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.26494611499, "elapsed_seconds": 11753.806013964, "env_steps_per_second": 4083.7835798016445, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1766307433266667, "eta_seconds": 7835.870675976, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.006303895518080083, "chaser_grad_norm": 0.36282528191804886, "chaser_new_fields": 59.875, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 15.319999694824219, "chaser_subspace_norm": 213.62905883789062, "collisions_per_episode": 12.75, "entropy": 0.13149448484182358, "explorer_grad_norm": 1.426378458738327, "explorer_new_fields": 71.3499984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 44.98749923706055, "explorer_subspace_norm": 85.52222442626953, "final_distance": 1.5731996297836304, "max_param_abs": 2.2504642009735107, "policy_loss": -0.0048599378122133885, "value_loss": 2.1862992644309998}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:14:01.944210+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034041.9442236, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.402270067648626, "elapsed_seconds": 12248.172243535053, "env_steps_per_second": 4082.2417423458, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.0413620405891755, "eta_seconds": 7348.903346121032, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0032204400656379306, "chaser_grad_norm": 0.8049173578619957, "chaser_new_fields": 59.04999923706055, "chaser_partner_vision": 0.9602500200271606, "chaser_return": 13.449999809265137, "chaser_subspace_norm": 204.0418701171875, "collisions_per_episode": 11.25, "entropy": 0.12901391088962555, "explorer_grad_norm": 2.153268739581108, "explorer_new_fields": 69.70000457763672, "explorer_partner_vision": 0.9602500200271606, "explorer_return": 43.525001525878906, "explorer_subspace_norm": 78.39869689941406, "final_distance": 1.9729379415512085, "max_param_abs": 2.2674977779388428, "policy_loss": -0.004101490026805266, "value_loss": 2.041717290878296}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:22:16.310439+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034536.310451, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5354413301658414, "elapsed_seconds": 12727.58878859703, "env_steps_per_second": 4085.6128260985397, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.903699177781607, "eta_seconds": 6853.317040013785, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.005255171789453783, "chaser_grad_norm": 0.7337998449802399, "chaser_new_fields": 53.22500228881836, "chaser_partner_vision": 0.95250004529953, "chaser_return": 16.410001754760742, "chaser_subspace_norm": 183.99581909179688, "collisions_per_episode": 14.949999809265137, "entropy": 0.19929227605462074, "explorer_grad_norm": 1.4212696403265, "explorer_new_fields": 60.875, "explorer_partner_vision": 0.95250004529953, "explorer_return": 29.5625, "explorer_subspace_norm": 68.93970489501953, "final_distance": 1.5210261344909668, "max_param_abs": 2.2965424060821533, "policy_loss": -0.0034024578491198554, "value_loss": 3.5272399187088013}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:30:15.726985+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035015.7269983, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.671206325868326, "elapsed_seconds": 13216.342773125973, "env_steps_per_second": 4085.8504449357397, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.767617860603268, "eta_seconds": 6363.424298171764, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0037371738480374006, "chaser_grad_norm": 0.7217032164335251, "chaser_new_fields": 54.92499923706055, "chaser_partner_vision": 0.9587500691413879, "chaser_return": 11.605000495910645, "chaser_subspace_norm": 188.16514587402344, "collisions_per_episode": 10.25, "entropy": 0.19815200194716454, "explorer_grad_norm": 1.3868125826120377, "explorer_new_fields": 62.875, "explorer_partner_vision": 0.9587500691413879, "explorer_return": 35.0625, "explorer_subspace_norm": 68.17832946777344, "final_distance": 1.557106852531433, "max_param_abs": 2.329324245452881, "policy_loss": -0.0047610752102968945, "value_loss": 2.759183347225189}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:38:24.480970+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035504.4809842, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.8084131099938854, "elapsed_seconds": 13710.287195977988, "env_steps_per_second": 4084.5242116028035, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6321770471402366, "eta_seconds": 5875.837369704852, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.004107472752663099, "chaser_grad_norm": 0.4520687833428383, "chaser_new_fields": 53.32500076293945, "chaser_partner_vision": 0.9660000205039978, "chaser_return": 17.344999313354492, "chaser_subspace_norm": 216.10476684570312, "collisions_per_episode": 15.75, "entropy": 0.1382647641003132, "explorer_grad_norm": 1.1971846520900726, "explorer_new_fields": 59.20000076293945, "explorer_partner_vision": 0.9660000205039978, "explorer_return": 27.475000381469727, "explorer_subspace_norm": 92.25552368164062, "final_distance": 1.5904531478881836, "max_param_abs": 2.353010416030884, "policy_loss": -0.002196004913429306, "value_loss": 3.5581236481666565}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:46:38.425393+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035998.4254076, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9454444890877736, "elapsed_seconds": 14203.600160715985, "env_steps_per_second": 4083.471749677604, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.496547909653983, "eta_seconds": 5387.572474754339, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.00416395960556315, "chaser_grad_norm": 0.6329823955893517, "chaser_new_fields": 55.04999923706055, "chaser_partner_vision": 0.9690000414848328, "chaser_return": 13.3725004196167, "chaser_subspace_norm": 223.11358642578125, "collisions_per_episode": 11.875, "entropy": 0.17521857097744942, "explorer_grad_norm": 1.15126733481884, "explorer_new_fields": 63.17499923706055, "explorer_partner_vision": 0.9690000414848328, "explorer_return": 34.775001525878906, "explorer_subspace_norm": 90.66083526611328, "final_distance": 1.559154748916626, "max_param_abs": 2.382112741470337, "policy_loss": -0.00495302220525673, "value_loss": 3.1336748600006104}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:54:51.738356+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036491.738367, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.0848425797508225, "elapsed_seconds": 14705.433287102962, "env_steps_per_second": 4080.1245926307743, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3616141932502743, "eta_seconds": 4901.811095700988, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0031635634384752365, "chaser_grad_norm": 0.5903649926185608, "chaser_new_fields": 58.775001525878906, "chaser_partner_vision": 0.9535000324249268, "chaser_return": 12.760001182556152, "chaser_subspace_norm": 238.91549682617188, "collisions_per_episode": 10.600000381469727, "entropy": 0.13946427032351494, "explorer_grad_norm": 1.0264621526002884, "explorer_new_fields": 67.70000457763672, "explorer_partner_vision": 0.9535000324249268, "explorer_return": 42.650001525878906, "explorer_subspace_norm": 82.36676788330078, "final_distance": 1.6521885395050049, "max_param_abs": 2.3992857933044434, "policy_loss": -0.0034682695004164543, "value_loss": 2.4903752207756042}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:03:13.571483+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036993.5714977, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.221260684896131, "elapsed_seconds": 15196.538465626072, "env_steps_per_second": 4079.8764889939494, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2255272956150058, "eta_seconds": 4411.898264214021, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0028482885377041356, "chaser_grad_norm": 0.5746633112430573, "chaser_new_fields": 60.97500228881836, "chaser_partner_vision": 0.9490000605583191, "chaser_return": 11.300000190734863, "chaser_subspace_norm": 251.2641143798828, "collisions_per_episode": 8.90000057220459, "entropy": 0.11875741183757782, "explorer_grad_norm": 1.5110651105642319, "explorer_new_fields": 71.9000015258789, "explorer_partner_vision": 0.9490000605583191, "explorer_return": 49.01250076293945, "explorer_subspace_norm": 98.71589660644531, "final_distance": 1.853376030921936, "max_param_abs": 2.4405620098114014, "policy_loss": -0.0043809087136743585, "value_loss": 2.217275083065033}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:11:24.676662+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037484.6766756, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.359916449337518, "elapsed_seconds": 15695.699217615067, "env_steps_per_second": 4077.5501054565116, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0899791123343796, "eta_seconds": 3923.9248044037668, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00416270857118306, "chaser_grad_norm": 0.3589393123984337, "chaser_new_fields": 56.35000228881836, "chaser_partner_vision": 0.9380000233650208, "chaser_return": 11.350000381469727, "chaser_subspace_norm": 244.2346954345703, "collisions_per_episode": 9.75, "entropy": 0.17046286165714264, "explorer_grad_norm": 0.9043444097042084, "explorer_new_fields": 65.75, "explorer_partner_vision": 0.9380000233650208, "explorer_return": 40.07500076293945, "explorer_subspace_norm": 89.46070098876953, "final_distance": 1.6945098638534546, "max_param_abs": 2.4677939414978027, "policy_loss": -0.005437442854969277, "value_loss": 3.041357934474945}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:19:43.837413+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037983.8374293, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.4934415737086155, "elapsed_seconds": 16176.389665351016, "env_steps_per_second": 4080.020410324843, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9531542732109185, "eta_seconds": 3431.3553835593066, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.003742935262721403, "chaser_grad_norm": 1.203482061624527, "chaser_new_fields": 56.82500076293945, "chaser_partner_vision": 0.956000030040741, "chaser_return": 11.452499389648438, "chaser_subspace_norm": 239.6360626220703, "collisions_per_episode": 9.77500057220459, "entropy": 0.13820801675319672, "explorer_grad_norm": 0.8169179856777191, "explorer_new_fields": 65.7249984741211, "explorer_partner_vision": 0.956000030040741, "explorer_return": 40.0625, "explorer_subspace_norm": 99.83335876464844, "final_distance": 1.7381728887557983, "max_param_abs": 2.5007567405700684, "policy_loss": -0.005906211157338248, "value_loss": 2.7475466132164}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:27:44.527861+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038464.527873, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.627833465080572, "elapsed_seconds": 16660.20047429006, "env_steps_per_second": 4081.583538261575, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8166764938377481, "eta_seconds": 2940.0353778158933, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0039545309992190525, "chaser_grad_norm": 0.46526166051626205, "chaser_new_fields": 58.775001525878906, "chaser_partner_vision": 0.9467500448226929, "chaser_return": 12.49000072479248, "chaser_subspace_norm": 243.65447998046875, "collisions_per_episode": 10.350000381469727, "entropy": 0.1382831484079361, "explorer_grad_norm": 1.2658189535140991, "explorer_new_fields": 64.2249984741211, "explorer_partner_vision": 0.9467500448226929, "explorer_return": 38.20000076293945, "explorer_subspace_norm": 98.98873138427734, "final_distance": 1.7356990575790405, "max_param_abs": 2.519597053527832, "policy_loss": -0.0044769548401148995, "value_loss": 2.8514263033866882}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:35:48.338669+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038948.3386815, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.761172588384458, "elapsed_seconds": 17140.22131818405, "env_steps_per_second": 4083.961268676096, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6801675126263511, "eta_seconds": 2448.603045454864, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0035778953306220274, "chaser_grad_norm": 0.3220148980617523, "chaser_new_fields": 59.60000228881836, "chaser_partner_vision": 0.936500072479248, "chaser_return": 11.005000114440918, "chaser_subspace_norm": 239.59046936035156, "collisions_per_episode": 8.800000190734863, "entropy": 0.15530100092291832, "explorer_grad_norm": 1.625396877527237, "explorer_new_fields": 68.17500305175781, "explorer_partner_vision": 0.936500072479248, "explorer_return": 44.337501525878906, "explorer_subspace_norm": 98.55020904541016, "final_distance": 1.6068744659423828, "max_param_abs": 2.5467963218688965, "policy_loss": -0.002934377118009479, "value_loss": 2.538341462612152}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:43:48.359513+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039428.3595247, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.897166290088337, "elapsed_seconds": 17629.798644318013, "env_steps_per_second": 4083.9944603227336, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5441295877875929, "eta_seconds": 1958.8665160353346, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00812048246711683, "chaser_grad_norm": 0.5164786502718925, "chaser_new_fields": 51.04999923706055, "chaser_partner_vision": 0.9602500200271606, "chaser_return": 15.370001792907715, "chaser_subspace_norm": 234.02078247070312, "collisions_per_episode": 14.300000190734863, "entropy": 0.1921299323439598, "explorer_grad_norm": 1.0539137870073318, "explorer_new_fields": 55.60000228881836, "explorer_partner_vision": 0.9602500200271606, "explorer_return": 23.100000381469727, "explorer_subspace_norm": 101.78244018554688, "final_distance": 1.8225024938583374, "max_param_abs": 2.559715509414673, "policy_loss": -0.004634681571303645, "value_loss": 3.9837526679039}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:51:57.936841+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039917.936856, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.03355571988278, "elapsed_seconds": 18120.80059157801, "env_steps_per_second": 4083.7047803722826, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4081261394499551, "eta_seconds": 1469.2541020198385, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.0024120734659476284, "chaser_grad_norm": 0.2763953134417534, "chaser_new_fields": 57.42499923706055, "chaser_partner_vision": 0.9592500329017639, "chaser_return": 12.572501182556152, "chaser_subspace_norm": 255.4188690185547, "collisions_per_episode": 10.725000381469727, "entropy": 0.1491260565817356, "explorer_grad_norm": 0.984371230006218, "explorer_new_fields": 66.25, "explorer_partner_vision": 0.9592500329017639, "explorer_return": 40.67499923706055, "explorer_subspace_norm": 107.37866973876953, "final_distance": 1.6753437519073486, "max_param_abs": 2.601377248764038, "policy_loss": -0.004898253655994722, "value_loss": 2.8549439907073975}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:00:08.938786+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040408.9387991, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.170164311979445, "elapsed_seconds": 18612.591523126, "env_steps_per_second": 4083.257288785959, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2721139111568129, "eta_seconds": 979.6100801645264, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.005293058960759822, "chaser_grad_norm": 0.4384599104523659, "chaser_new_fields": 56.29999923706055, "chaser_partner_vision": 0.968250036239624, "chaser_return": 16.670001983642578, "chaser_subspace_norm": 275.66021728515625, "collisions_per_episode": 14.75, "entropy": 0.1576538383960724, "explorer_grad_norm": 0.8651386946439743, "explorer_new_fields": 66.0999984741211, "explorer_partner_vision": 0.968250036239624, "explorer_return": 36.78750228881836, "explorer_subspace_norm": 115.39480590820312, "final_distance": 1.8006147146224976, "max_param_abs": 2.652438163757324, "policy_loss": -0.004232267527413569, "value_loss": 3.4680804014205933}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:08:20.729721+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040900.7297375, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.303483881565852, "elapsed_seconds": 19092.541973637068, "env_steps_per_second": 4085.3648564817718, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.1359867661939962, "eta_seconds": 489.5523582983863, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.005887299084726294, "chaser_grad_norm": 0.5030369907617569, "chaser_new_fields": 49.97500228881836, "chaser_partner_vision": 0.9532500505447388, "chaser_return": 18.07000160217285, "chaser_subspace_norm": 253.5067596435547, "collisions_per_episode": 17.149999618530273, "entropy": 0.16699428856372833, "explorer_grad_norm": 0.7504595667123795, "explorer_new_fields": 57.75, "explorer_partner_vision": 0.9532500505447388, "explorer_return": 23.100000381469727, "explorer_subspace_norm": 113.88228607177734, "final_distance": 1.3508764505386353, "max_param_abs": 2.6431610584259033, "policy_loss": -0.004097296256581728, "value_loss": 3.947905123233795}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:16:20.680168+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041380.6801805, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.425114343349996, "elapsed_seconds": 19530.411636059987, "env_steps_per_second": 4096.175825208515, "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-transformer/social/seed_0007/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:38.549830+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041818.5498412, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.425122272546384, "elapsed_seconds": 19530.44018116698, "env_steps_per_second": 4096.169838360492, "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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.004598107310812227, "chaser_grad_norm": 0.545984223484993, "chaser_new_fields": 57.54999923706055, "chaser_partner_vision": 0.9745000600814819, "chaser_return": 13.77500057220459, "chaser_subspace_norm": 288.3200988769531, "collisions_per_episode": 11.90000057220459, "entropy": 0.1392931044101715, "explorer_grad_norm": 0.9233212769031525, "explorer_new_fields": 65.2249984741211, "explorer_partner_vision": 0.9745000600814819, "explorer_return": 38.17499923706055, "explorer_subspace_norm": 111.23908996582031, "final_distance": 1.6038007736206055, "max_param_abs": 2.665212392807007, "policy_loss": -0.005061323422069108, "value_loss": 3.0026434659957886}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:38.578376+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041818.5783887, "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..56cb798dd5129b239b00e48927fc2be82ad34ca7 --- /dev/null +++ b/social/seed_0007/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.425122272546384, + "elapsed_seconds": 19530.44018116698, + "env_steps_per_second": 4096.169838360492, + "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-transformer/social/seed_0007/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.004598107310812227, + "chaser_grad_norm": 0.545984223484993, + "chaser_new_fields": 57.54999923706055, + "chaser_partner_vision": 0.9745000600814819, + "chaser_return": 13.77500057220459, + "chaser_subspace_norm": 288.3200988769531, + "collisions_per_episode": 11.90000057220459, + "entropy": 0.1392931044101715, + "explorer_grad_norm": 0.9233212769031525, + "explorer_new_fields": 65.2249984741211, + "explorer_partner_vision": 0.9745000600814819, + "explorer_return": 38.17499923706055, + "explorer_subspace_norm": 111.23908996582031, + "final_distance": 1.6038007736206055, + "max_param_abs": 2.665212392807007, + "policy_loss": -0.005061323422069108, + "value_loss": 3.0026434659957886 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:38.578376+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041818.5783887, + "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..e693354e48d1b5506fcfbde77a650b6475e6815c --- /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:ff3f0e0f69d3eb842d30e1463bfb424b5f279fa52d10afb3cf854b7b57747b13 +size 29760384 diff --git a/social/seed_0008/attempt_01/config.json b/social/seed_0008/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..c919c294344bf171abe411af4220c1ac277b5994 --- /dev/null +++ b/social/seed_0008/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0008/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..4de0bf299f0d9059ed28b92abbb20e9fc2fe2221 --- /dev/null +++ b/social/seed_0008/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.0007043890152837978, "elapsed_seconds": 2.535800455021672, "env_steps_per_second": 1577.411184732126, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.087075916660671, "eta_seconds": 50713.47329997842, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004260548553778776, "chaser_grad_norm": 4.733773998916149, "chaser_new_fields": 34.75, "chaser_partner_vision": 0.23525001108646393, "chaser_return": -2.2375004291534424, "chaser_subspace_norm": 0.04183848947286606, "collisions_per_episode": 0.7750000357627869, "entropy": 2.7514923810958862, "explorer_grad_norm": 6.185827016830444, "explorer_new_fields": 33.525001525878906, "explorer_partner_vision": 0.23525001108646393, "explorer_return": -0.2874999940395355, "explorer_subspace_norm": 0.036795929074287415, "final_distance": 6.356278896331787, "max_param_abs": 1.001199722290039, "policy_loss": -0.014282301004527653, "value_loss": 8.240208983421326}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:58:20.796184+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022300.7962, "update": 1} +{"device": "cuda", "elapsed_hours": 0.13410934275084097, "elapsed_seconds": 482.7936339030275, "env_steps_per_second": 4142.556694112735, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.230264367282798, "eta_seconds": 18828.951722218073, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0028492887640112485, "chaser_grad_norm": 0.9594683796167374, "chaser_new_fields": 60.35000228881836, "chaser_partner_vision": 0.4592500329017639, "chaser_return": 14.225000381469727, "chaser_subspace_norm": 9.7769775390625, "collisions_per_episode": 11.300000190734863, "entropy": 0.5916979014873505, "explorer_grad_norm": 1.8573675751686096, "explorer_new_fields": 64.17500305175781, "explorer_partner_vision": 0.4592500329017639, "explorer_return": 37.462501525878906, "explorer_subspace_norm": 6.615767478942871, "final_distance": 4.986347198486328, "max_param_abs": 1.162577748298645, "policy_loss": -0.00532318285666733, "value_loss": 3.7106704711914062}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:06:21.054018+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022781.0540366, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2706081499611092, "elapsed_seconds": 974.1893398599932, "env_steps_per_second": 4105.978002771889, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.141554849261075, "eta_seconds": 18509.59745733987, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.002860739614858776, "chaser_grad_norm": 1.6783881485462189, "chaser_new_fields": 57.82500076293945, "chaser_partner_vision": 0.4982500374317169, "chaser_return": 12.02500057220459, "chaser_subspace_norm": 16.363555908203125, "collisions_per_episode": 9.800000190734863, "entropy": 0.46951188892126083, "explorer_grad_norm": 1.7360083162784576, "explorer_new_fields": 67.375, "explorer_partner_vision": 0.4982500374317169, "explorer_return": 42.375, "explorer_subspace_norm": 11.8851900100708, "final_distance": 3.984112501144409, "max_param_abs": 1.2923400402069092, "policy_loss": -0.005909473638979534, "value_loss": 3.350748360157013}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:14:32.449723+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023272.449737, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.4087821827319567, "elapsed_seconds": 1471.6158578350442, "env_steps_per_second": 4077.1509548876784, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.041646920360799, "eta_seconds": 18149.928913298878, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00300406188810598, "chaser_grad_norm": 0.5042952820658684, "chaser_new_fields": 58.70000076293945, "chaser_partner_vision": 0.4987500309944153, "chaser_return": 10.83750057220459, "chaser_subspace_norm": 20.817636489868164, "collisions_per_episode": 8.52500057220459, "entropy": 0.46614885330200195, "explorer_grad_norm": 1.597964733839035, "explorer_new_fields": 67.5, "explorer_partner_vision": 0.4987500309944153, "explorer_return": 44.025001525878906, "explorer_subspace_norm": 15.751598358154297, "final_distance": 3.50532603263855, "max_param_abs": 1.4205254316329956, "policy_loss": -0.005943751469700942, "value_loss": 3.020463526248932}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:22:49.876239+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023769.87625, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5432215113024964, "elapsed_seconds": 1955.5974406889873, "env_steps_per_second": 4090.821471509738, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.888993601722468, "eta_seconds": 17600.376966200885, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0050867291023379835, "chaser_grad_norm": 0.7601244300603867, "chaser_new_fields": 52.32500076293945, "chaser_partner_vision": 0.5847500562667847, "chaser_return": 12.690000534057617, "chaser_subspace_norm": 23.581554412841797, "collisions_per_episode": 11.449999809265137, "entropy": 0.441886305809021, "explorer_grad_norm": 1.8597585558891296, "explorer_new_fields": 65.375, "explorer_partner_vision": 0.5847500562667847, "explorer_return": 38.962501525878906, "explorer_subspace_norm": 21.836864471435547, "final_distance": 3.440647602081299, "max_param_abs": 1.5176725387573242, "policy_loss": -0.007126769926401266, "value_loss": 3.501602053642273}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:30:53.857824+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024253.8578398, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6781592971163789, "elapsed_seconds": 2441.373469618964, "env_steps_per_second": 4096.054997091757, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.747115079814653, "eta_seconds": 17089.614287332748, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.0032572916069497637, "chaser_grad_norm": 0.6438449770212173, "chaser_new_fields": 51.82500076293945, "chaser_partner_vision": 0.6107500195503235, "chaser_return": 10.977500915527344, "chaser_subspace_norm": 24.420927047729492, "collisions_per_episode": 9.975000381469727, "entropy": 0.4972897469997406, "explorer_grad_norm": 1.9603288769721985, "explorer_new_fields": 65.2750015258789, "explorer_partner_vision": 0.6107500195503235, "explorer_return": 40.1875, "explorer_subspace_norm": 22.8709659576416, "final_distance": 3.518876314163208, "max_param_abs": 1.5701255798339844, "policy_loss": -0.009037294078616753, "value_loss": 3.4363311529159546}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:38:59.633850+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024739.6338627, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8116724911230591, "elapsed_seconds": 2922.0209680430125, "env_steps_per_second": 4106.746711005586, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.599477449697334, "eta_seconds": 16558.118818910403, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.005900444522392689, "chaser_grad_norm": 1.0091767609119415, "chaser_new_fields": 44.125, "chaser_partner_vision": 0.7440000176429749, "chaser_return": 15.945001602172852, "chaser_subspace_norm": 27.81255340576172, "collisions_per_episode": 16.0, "entropy": 0.46081550419330597, "explorer_grad_norm": 1.7298774123191833, "explorer_new_fields": 61.35000228881836, "explorer_partner_vision": 0.7440000176429749, "explorer_return": 29.787500381469727, "explorer_subspace_norm": 23.108261108398438, "final_distance": 2.3980252742767334, "max_param_abs": 1.6368566751480103, "policy_loss": -0.009449649900204804, "value_loss": 4.24945855140686}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:47:00.281351+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025220.281366, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9486350156699902, "elapsed_seconds": 3415.086056411965, "env_steps_per_second": 4099.457456925404, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.47213650244424, "eta_seconds": 16099.691408799265, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.002603997279926773, "chaser_grad_norm": 0.822179950773716, "chaser_new_fields": 50.29999923706055, "chaser_partner_vision": 0.9177500605583191, "chaser_return": 20.360000610351562, "chaser_subspace_norm": 32.49226379394531, "collisions_per_episode": 19.200000762939453, "entropy": 0.38377803564071655, "explorer_grad_norm": 1.8995552957057953, "explorer_new_fields": 61.47500228881836, "explorer_partner_vision": 0.9177500605583191, "explorer_return": 25.5625, "explorer_subspace_norm": 25.924123764038086, "final_distance": 2.355350971221924, "max_param_abs": 1.6594164371490479, "policy_loss": -0.007138265739229155, "value_loss": 3.505883038043976}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:55:13.346439+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025713.3464525, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.082724426222784, "elapsed_seconds": 3897.8079344020225, "env_steps_per_second": 4104.871319795961, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.330897704891136, "eta_seconds": 15591.23173760809, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.004441044340779854, "chaser_grad_norm": 0.6165179014205933, "chaser_new_fields": 51.5, "chaser_partner_vision": 0.9485000371932983, "chaser_return": 17.07750129699707, "chaser_subspace_norm": 50.79289627075195, "collisions_per_episode": 15.675000190734863, "entropy": 0.2843433916568756, "explorer_grad_norm": 4.107379913330078, "explorer_new_fields": 61.35000228881836, "explorer_partner_vision": 0.9485000371932983, "explorer_return": 28.600000381469727, "explorer_subspace_norm": 32.2357292175293, "final_distance": 2.0911662578582764, "max_param_abs": 1.6932071447372437, "policy_loss": -0.0062115670954767666, "value_loss": 3.678258001804352}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:03:16.068316+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026196.0683303, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2168149031275082, "elapsed_seconds": 4380.533651259029, "env_steps_per_second": 4109.08839721538, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.19125133299475, "eta_seconds": 15088.504798781101, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.0061858909701904, "chaser_grad_norm": 0.45489102602005005, "chaser_new_fields": 48.92499923706055, "chaser_partner_vision": 0.9660000205039978, "chaser_return": 19.107501983642578, "chaser_subspace_norm": 54.497039794921875, "collisions_per_episode": 18.17500114440918, "entropy": 0.2806348130106926, "explorer_grad_norm": 1.5437028408050537, "explorer_new_fields": 57.92499923706055, "explorer_partner_vision": 0.9660000205039978, "explorer_return": 21.575000762939453, "explorer_subspace_norm": 31.947391510009766, "final_distance": 1.978772759437561, "max_param_abs": 1.7188080549240112, "policy_loss": -0.004474085962399954, "value_loss": 3.7602802515029907}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:11:18.794032+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026678.794044, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.3519056930483202, "elapsed_seconds": 4866.860494973953, "env_steps_per_second": 4109.425371993745, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.055717079144961, "eta_seconds": 14600.58148492186, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.006289718451333748, "chaser_grad_norm": 1.4274334013462067, "chaser_new_fields": 45.97500228881836, "chaser_partner_vision": 0.9615000486373901, "chaser_return": 23.767501831054688, "chaser_subspace_norm": 61.78669738769531, "collisions_per_episode": 23.024999618530273, "entropy": 0.27741552889347076, "explorer_grad_norm": 1.337012991309166, "explorer_new_fields": 50.22500228881836, "explorer_partner_vision": 0.9615000486373901, "explorer_return": 8.387499809265137, "explorer_subspace_norm": 33.72709274291992, "final_distance": 1.734770655632019, "max_param_abs": 1.7597085237503052, "policy_loss": -0.004716003681531689, "value_loss": 4.7889991998672485}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:19:25.120876+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027165.1208918, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.489037105878039, "elapsed_seconds": 5360.53358116094, "env_steps_per_second": 4104.069057102226, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.925643279133012, "eta_seconds": 14132.315804878843, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.0045229399773560885, "chaser_grad_norm": 0.5726823657751083, "chaser_new_fields": 51.60000228881836, "chaser_partner_vision": 0.9495000243186951, "chaser_return": 14.227499961853027, "chaser_subspace_norm": 64.88787078857422, "collisions_per_episode": 13.225000381469727, "entropy": 0.2579587996006012, "explorer_grad_norm": 1.4241363108158112, "explorer_new_fields": 60.20000076293945, "explorer_partner_vision": 0.9495000243186951, "explorer_return": 28.962499618530273, "explorer_subspace_norm": 34.75232696533203, "final_distance": 1.6443456411361694, "max_param_abs": 1.8433034420013428, "policy_loss": -0.0069715264458238835, "value_loss": 3.6189242601394653}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:27:38.793965+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027658.7939792, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.62770288109055, "elapsed_seconds": 5859.73037192598, "env_steps_per_second": 4095.7515920842047, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.797973389211283, "eta_seconds": 13672.704201160619, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.0032478380467999557, "chaser_grad_norm": 0.5502270609140396, "chaser_new_fields": 54.17499923706055, "chaser_partner_vision": 0.9660000205039978, "chaser_return": 14.8725004196167, "chaser_subspace_norm": 74.77294921875, "collisions_per_episode": 13.475000381469727, "entropy": 0.23484285920858383, "explorer_grad_norm": 1.2442690879106522, "explorer_new_fields": 60.25, "explorer_partner_vision": 0.9660000205039978, "explorer_return": 29.212499618530273, "explorer_subspace_norm": 37.214168548583984, "final_distance": 1.7999470233917236, "max_param_abs": 1.9321426153182983, "policy_loss": -0.006528030651622885, "value_loss": 3.1174720525741577}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:35:57.990755+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028157.9907703, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7644432576197302, "elapsed_seconds": 6351.995727431029, "env_steps_per_second": 4093.201745668572, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6646129196717476, "eta_seconds": 13192.606510818292, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.003300463902991879, "chaser_grad_norm": 0.3933399096131325, "chaser_new_fields": 54.900001525878906, "chaser_partner_vision": 0.9562500715255737, "chaser_return": 13.692500114440918, "chaser_subspace_norm": 78.43421936035156, "collisions_per_episode": 12.175000190734863, "entropy": 0.2737230286002159, "explorer_grad_norm": 1.4712141454219818, "explorer_new_fields": 62.57500076293945, "explorer_partner_vision": 0.9562500715255737, "explorer_return": 33.79999923706055, "explorer_subspace_norm": 36.061607360839844, "final_distance": 1.7711613178253174, "max_param_abs": 1.9848028421401978, "policy_loss": -0.0063707898724825185, "value_loss": 3.2417526245117188}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:44:10.256108+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028650.2561195, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.8976968841513735, "elapsed_seconds": 6831.708782944945, "env_steps_per_second": 4098.53535764006, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.5242942134239796, "eta_seconds": 12687.459168326326, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004073718935149628, "chaser_grad_norm": 0.4040597975254059, "chaser_new_fields": 57.20000076293945, "chaser_partner_vision": 0.968000054359436, "chaser_return": 12.67750072479248, "chaser_subspace_norm": 88.43878936767578, "collisions_per_episode": 10.77500057220459, "entropy": 0.22373629733920097, "explorer_grad_norm": 1.2737500667572021, "explorer_new_fields": 62.42499923706055, "explorer_partner_vision": 0.968000054359436, "explorer_return": 34.72500228881836, "explorer_subspace_norm": 42.15757751464844, "final_distance": 1.7975704669952393, "max_param_abs": 2.0267231464385986, "policy_loss": -0.006358268270261647, "value_loss": 3.32760751247406}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:52:09.969164+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029129.969179, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.032531192304159, "elapsed_seconds": 7317.112292294973, "env_steps_per_second": 4099.978079001254, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.3875519871735986, "eta_seconds": 12195.187153824954, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.0038897674473314936, "chaser_grad_norm": 0.633298896253109, "chaser_new_fields": 54.275001525878906, "chaser_partner_vision": 0.971500039100647, "chaser_return": 13.125, "chaser_subspace_norm": 101.55253601074219, "collisions_per_episode": 11.850000381469727, "entropy": 0.20207776874303818, "explorer_grad_norm": 1.5442892909049988, "explorer_new_fields": 61.22500228881836, "explorer_partner_vision": 0.971500039100647, "explorer_return": 31.86250114440918, "explorer_subspace_norm": 42.437156677246094, "final_distance": 2.0445797443389893, "max_param_abs": 2.0815305709838867, "policy_loss": -0.006261134067528928, "value_loss": 2.9122445583343506}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:00:15.372673+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029615.3726845, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.170663557260543, "elapsed_seconds": 7814.388806137955, "env_steps_per_second": 4095.009960966495, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.2559953358908147, "eta_seconds": 11721.583209206932, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0035139690748238195, "chaser_grad_norm": 0.6568476855754852, "chaser_new_fields": 53.32500076293945, "chaser_partner_vision": 0.9612500667572021, "chaser_return": 14.222500801086426, "chaser_subspace_norm": 101.65321350097656, "collisions_per_episode": 13.02500057220459, "entropy": 0.22066842019557953, "explorer_grad_norm": 1.5047696828842163, "explorer_new_fields": 63.10000228881836, "explorer_partner_vision": 0.9612500667572021, "explorer_return": 33.41250228881836, "explorer_subspace_norm": 48.071990966796875, "final_distance": 1.9562841653823853, "max_param_abs": 2.070988655090332, "policy_loss": -0.006652241269122783, "value_loss": 3.3188514709472656}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:08:32.649189+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030112.649205, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.306803488078618, "elapsed_seconds": 8304.492557083024, "env_steps_per_second": 4094.1694831192185, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1209694250475417, "eta_seconds": 11235.48993017115, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004679206011443782, "chaser_grad_norm": 0.6125284135341644, "chaser_new_fields": 55.525001525878906, "chaser_partner_vision": 0.952750027179718, "chaser_return": 15.592500686645508, "chaser_subspace_norm": 112.56909942626953, "collisions_per_episode": 13.675000190734863, "entropy": 0.2084084339439869, "explorer_grad_norm": 1.6461684703826904, "explorer_new_fields": 66.0250015258789, "explorer_partner_vision": 0.952750027179718, "explorer_return": 38.48749923706055, "explorer_subspace_norm": 48.9841194152832, "final_distance": 1.9756149053573608, "max_param_abs": 2.1265156269073486, "policy_loss": -0.004745281291675951, "value_loss": 2.9206555485725403}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:16:42.752939+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030602.7529495, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.443516326353064, "elapsed_seconds": 8796.65877487103, "env_steps_per_second": 4092.462936363904, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.9865199544315226, "eta_seconds": 10751.471835953482, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.0038717759545452424, "chaser_grad_norm": 0.5223696306347847, "chaser_new_fields": 54.775001525878906, "chaser_partner_vision": 0.9650000333786011, "chaser_return": 12.77500057220459, "chaser_subspace_norm": 111.87095642089844, "collisions_per_episode": 11.350000381469727, "entropy": 0.2160603478550911, "explorer_grad_norm": 1.1861671060323715, "explorer_new_fields": 62.17499923706055, "explorer_partner_vision": 0.9650000333786011, "explorer_return": 34.0625, "explorer_subspace_norm": 57.54012680053711, "final_distance": 1.8788397312164307, "max_param_abs": 2.1624152660369873, "policy_loss": -0.005958274887873927, "value_loss": 3.180261194705963}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:24:54.919158+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031094.919174, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.576963614681671, "elapsed_seconds": 9277.069012854015, "env_steps_per_second": 4096.121301603814, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8482229425428995, "eta_seconds": 10253.60259315444, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.004734078174182699, "chaser_grad_norm": 0.7114104181528091, "chaser_new_fields": 56.900001525878906, "chaser_partner_vision": 0.9585000276565552, "chaser_return": 11.720000267028809, "chaser_subspace_norm": 121.18279266357422, "collisions_per_episode": 9.949999809265137, "entropy": 0.20013543218374252, "explorer_grad_norm": 0.8736875802278519, "explorer_new_fields": 63.625, "explorer_partner_vision": 0.9585000276565552, "explorer_return": 37.91250228881836, "explorer_subspace_norm": 54.72572326660156, "final_distance": 1.76142156124115, "max_param_abs": 2.1841964721679688, "policy_loss": -0.006205531636283013, "value_loss": 2.9111461639404297}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:32:55.329394+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031575.3294058, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.714380684534715, "elapsed_seconds": 9771.770464324974, "env_steps_per_second": 4093.42402648865, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.714380684534715, "eta_seconds": 9771.770464324974, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.004449226981549259, "chaser_grad_norm": 0.5562428683042526, "chaser_new_fields": 58.29999923706055, "chaser_partner_vision": 0.9532500505447388, "chaser_return": 11.887500762939453, "chaser_subspace_norm": 116.83595275878906, "collisions_per_episode": 9.925000190734863, "entropy": 0.2064518742263317, "explorer_grad_norm": 0.9571553021669388, "explorer_new_fields": 67.5, "explorer_partner_vision": 0.9532500505447388, "explorer_return": 41.712501525878906, "explorer_subspace_norm": 54.71220397949219, "final_distance": 1.7727348804473877, "max_param_abs": 2.203671455383301, "policy_loss": -0.007127812724558424, "value_loss": 2.6032004952430725}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:41:10.030846+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032070.0308614, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8504979379530413, "elapsed_seconds": 10261.792576630949, "env_steps_per_second": 4092.8521685037827, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.579021943862276, "eta_seconds": 9284.478997904193, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.003660186701015533, "chaser_grad_norm": 0.8461195528507233, "chaser_new_fields": 57.72500228881836, "chaser_partner_vision": 0.9652500748634338, "chaser_return": 12.27500057220459, "chaser_subspace_norm": 133.40077209472656, "collisions_per_episode": 10.300000190734863, "entropy": 0.21211831271648407, "explorer_grad_norm": 1.2063899785280228, "explorer_new_fields": 65.0250015258789, "explorer_partner_vision": 0.9652500748634338, "explorer_return": 38.41250228881836, "explorer_subspace_norm": 53.89516067504883, "final_distance": 1.7776240110397339, "max_param_abs": 2.239967107772827, "policy_loss": -0.005671611065963589, "value_loss": 3.075587570667267}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:49:20.052958+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032560.0529692, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.985521320865276, "elapsed_seconds": 10747.876755114994, "env_steps_per_second": 4093.8318332558174, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4426992625261352, "eta_seconds": 8793.717345094086, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.003507472062741712, "chaser_grad_norm": 0.31608394905924797, "chaser_new_fields": 57.97500228881836, "chaser_partner_vision": 0.9565000534057617, "chaser_return": 11.8125, "chaser_subspace_norm": 138.87464904785156, "collisions_per_episode": 9.875, "entropy": 0.18797430023550987, "explorer_grad_norm": 1.0695826560258865, "explorer_new_fields": 66.9000015258789, "explorer_partner_vision": 0.9565000534057617, "explorer_return": 41.400001525878906, "explorer_subspace_norm": 60.44005584716797, "final_distance": 1.671122431755066, "max_param_abs": 2.2764036655426025, "policy_loss": -0.0051805924540389015, "value_loss": 2.680907905101776}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:57:26.137136+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033046.1371515, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1213413602486253, "elapsed_seconds": 11236.828896895051, "env_steps_per_second": 4093.6816269143933, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.307078396705506, "eta_seconds": 8305.482228139821, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.003595616509602273, "chaser_grad_norm": 1.0696356892585754, "chaser_new_fields": 49.79999923706055, "chaser_partner_vision": 0.9377500414848328, "chaser_return": 18.5, "chaser_subspace_norm": 124.98912048339844, "collisions_per_episode": 17.350000381469727, "entropy": 0.2848367393016815, "explorer_grad_norm": 1.123789668083191, "explorer_new_fields": 56.22500228881836, "explorer_partner_vision": 0.9377500414848328, "explorer_return": 22.36250114440918, "explorer_subspace_norm": 64.1504898071289, "final_distance": 2.0059595108032227, "max_param_abs": 2.3361334800720215, "policy_loss": -0.0034887302743698, "value_loss": 4.834597945213318}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:05:35.089279+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033535.0892901, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.257152524168616, "elapsed_seconds": 11725.749087007018, "env_steps_per_second": 4093.5551020094304, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.1714350161124107, "eta_seconds": 7817.166058004679, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.002512128558885607, "chaser_grad_norm": 0.6650192588567734, "chaser_new_fields": 56.525001525878906, "chaser_partner_vision": 0.9720000624656677, "chaser_return": 13.305002212524414, "chaser_subspace_norm": 137.60707092285156, "collisions_per_episode": 11.40000057220459, "entropy": 0.15298765152692795, "explorer_grad_norm": 1.008911356329918, "explorer_new_fields": 66.57500457763672, "explorer_partner_vision": 0.9720000624656677, "explorer_return": 40.1875, "explorer_subspace_norm": 69.0353775024414, "final_distance": 1.6978992223739624, "max_param_abs": 2.3831706047058105, "policy_loss": -0.004497652581889611, "value_loss": 2.944609582424164}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:13:44.009470+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034024.009484, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.3916058925750097, "elapsed_seconds": 12209.781213270035, "env_steps_per_second": 4095.0774732685777, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.034963535545006, "eta_seconds": 7325.868727962021, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.0030841273078658915, "chaser_grad_norm": 0.3096652254462242, "chaser_new_fields": 53.25, "chaser_partner_vision": 0.9610000252723694, "chaser_return": 15.21250057220459, "chaser_subspace_norm": 128.50527954101562, "collisions_per_episode": 13.824999809265137, "entropy": 0.21590346097946167, "explorer_grad_norm": 1.1189946085214615, "explorer_new_fields": 61.32500076293945, "explorer_partner_vision": 0.9610000252723694, "explorer_return": 31.325000762939453, "explorer_subspace_norm": 70.46833801269531, "final_distance": 1.8846193552017212, "max_param_abs": 2.384493827819824, "policy_loss": -0.00430736190964609, "value_loss": 3.5003146529197693}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:21:48.041594+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034508.041607, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5272280329847128, "elapsed_seconds": 12698.020918744965, "env_steps_per_second": 4095.126345495068, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8992766331456146, "eta_seconds": 6837.395879324213, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0059160173804002625, "chaser_grad_norm": 1.259853482246399, "chaser_new_fields": 51.60000228881836, "chaser_partner_vision": 0.9397500157356262, "chaser_return": 15.225000381469727, "chaser_subspace_norm": 181.7108612060547, "collisions_per_episode": 14.050000190734863, "entropy": 0.191739771515131, "explorer_grad_norm": 0.8253523260354996, "explorer_new_fields": 59.04999923706055, "explorer_partner_vision": 0.9397500157356262, "explorer_return": 27.274999618530273, "explorer_subspace_norm": 75.96083068847656, "final_distance": 1.670686960220337, "max_param_abs": 2.4197113513946533, "policy_loss": -0.004631245506928394, "value_loss": 3.737886130809784}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:29:56.281302+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034996.2813184, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.662739110419708, "elapsed_seconds": 13185.860797510948, "env_steps_per_second": 4095.2957739545836, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7635410531650446, "eta_seconds": 6348.74779139416, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0016599151997596095, "chaser_grad_norm": 0.4176120012998581, "chaser_new_fields": 55.95000076293945, "chaser_partner_vision": 0.9665000438690186, "chaser_return": 14.695001602172852, "chaser_subspace_norm": 182.65304565429688, "collisions_per_episode": 12.949999809265137, "entropy": 0.17899120226502419, "explorer_grad_norm": 1.0834126472473145, "explorer_new_fields": 64.17500305175781, "explorer_partner_vision": 0.9665000438690186, "explorer_return": 35.88750076293945, "explorer_subspace_norm": 68.7138442993164, "final_distance": 1.5774208307266235, "max_param_abs": 2.450813055038452, "policy_loss": -0.0026372367183000733, "value_loss": 3.2620392441749573}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:38:04.121179+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035484.121191, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.7962937374572117, "elapsed_seconds": 13666.657454845961, "env_steps_per_second": 4097.56373751384, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6269830303388046, "eta_seconds": 5857.138909219697, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0028335670419390446, "chaser_grad_norm": 0.6194358617067337, "chaser_new_fields": 58.60000228881836, "chaser_partner_vision": 0.9470000267028809, "chaser_return": 10.65000057220459, "chaser_subspace_norm": 163.0164794921875, "collisions_per_episode": 8.699999809265137, "entropy": 0.1959933154284954, "explorer_grad_norm": 1.8194412887096405, "explorer_new_fields": 67.625, "explorer_partner_vision": 0.9470000267028809, "explorer_return": 44.275001525878906, "explorer_subspace_norm": 70.04133605957031, "final_distance": 1.8990610837936401, "max_param_abs": 2.517655849456787, "policy_loss": -0.005496715434844646, "value_loss": 2.626596987247467}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:46:04.917837+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035964.9178493, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9297418586327693, "elapsed_seconds": 14147.07069107797, "env_steps_per_second": 4099.788660600843, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4905917394813955, "eta_seconds": 5366.130262133023, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0028686829927782043, "chaser_grad_norm": 0.7873056977987289, "chaser_new_fields": 58.625, "chaser_partner_vision": 0.9505000710487366, "chaser_return": 10.522500991821289, "chaser_subspace_norm": 185.95050048828125, "collisions_per_episode": 8.574999809265137, "entropy": 0.16639798507094383, "explorer_grad_norm": 0.9225423485040665, "explorer_new_fields": 69.625, "explorer_partner_vision": 0.9505000710487366, "explorer_return": 45.875, "explorer_subspace_norm": 78.40502166748047, "final_distance": 1.8867504596710205, "max_param_abs": 2.523698568344116, "policy_loss": -0.006137167414664191, "value_loss": 2.7081185579299927}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:54:05.331072+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036445.3310833, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.063236263275838, "elapsed_seconds": 14627.650547793019, "env_steps_per_second": 4101.820713036697, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.354412087758613, "eta_seconds": 4875.8835159310065, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.002574981807818233, "chaser_grad_norm": 0.6205654591321945, "chaser_new_fields": 57.67499923706055, "chaser_partner_vision": 0.968500018119812, "chaser_return": 9.512499809265137, "chaser_subspace_norm": 184.41952514648438, "collisions_per_episode": 7.724999904632568, "entropy": 0.16057820245623589, "explorer_grad_norm": 0.8486518263816833, "explorer_new_fields": 67.125, "explorer_partner_vision": 0.968500018119812, "explorer_return": 43.67499923706055, "explorer_subspace_norm": 79.87229919433594, "final_distance": 1.826462984085083, "max_param_abs": 2.575615406036377, "policy_loss": -0.005475931943408341, "value_loss": 2.532200574874878}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:02:05.910930+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036925.9109445, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.1967892202666714, "elapsed_seconds": 15108.441192960017, "env_steps_per_second": 4103.666235858253, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2184226768516144, "eta_seconds": 4386.321636665812, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0040895828681382085, "chaser_grad_norm": 0.4888424798846245, "chaser_new_fields": 58.54999923706055, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 11.039999961853027, "chaser_subspace_norm": 181.82424926757812, "collisions_per_episode": 9.0, "entropy": 0.16943693906068802, "explorer_grad_norm": 1.0799733251333237, "explorer_new_fields": 68.0999984741211, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 45.0625, "explorer_subspace_norm": 79.75861358642578, "final_distance": 2.0433201789855957, "max_param_abs": 2.593337059020996, "policy_loss": -0.005177549814806781, "value_loss": 2.2792734503746033}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:10:06.701574+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037406.7015846, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.330091263654726, "elapsed_seconds": 15588.328549157013, "env_steps_per_second": 4105.635815807911, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0825228159136813, "eta_seconds": 3897.082137289253, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.002418259655916799, "chaser_grad_norm": 0.9242428541183472, "chaser_new_fields": 57.60000228881836, "chaser_partner_vision": 0.9642500281333923, "chaser_return": 13.5600004196167, "chaser_subspace_norm": 193.14125061035156, "collisions_per_episode": 11.65000057220459, "entropy": 0.16423413902521133, "explorer_grad_norm": 1.54600590467453, "explorer_new_fields": 67.20000457763672, "explorer_partner_vision": 0.9642500281333923, "explorer_return": 41.29999923706055, "explorer_subspace_norm": 88.25200653076172, "final_distance": 1.8342227935791016, "max_param_abs": 2.6158030033111572, "policy_loss": -0.00554002732318537, "value_loss": 2.724867105484009}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:18:06.588931+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037886.5889473, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.4670924091580675, "elapsed_seconds": 16081.532672969042, "env_steps_per_second": 4104.086428959436, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9475650564880749, "eta_seconds": 3411.2342033570694, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0015222342821630264, "chaser_grad_norm": 0.574575237929821, "chaser_new_fields": 59.85000228881836, "chaser_partner_vision": 0.9597500562667847, "chaser_return": 11.25, "chaser_subspace_norm": 224.1264190673828, "collisions_per_episode": 9.0, "entropy": 0.1539052352309227, "explorer_grad_norm": 0.9615499079227448, "explorer_new_fields": 67.6500015258789, "explorer_partner_vision": 0.9597500562667847, "explorer_return": 43.5625, "explorer_subspace_norm": 87.2422866821289, "final_distance": 1.8510407209396362, "max_param_abs": 2.6452455520629883, "policy_loss": -0.005379022467877803, "value_loss": 2.264812171459198}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:26:19.793054+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038379.793066, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.605918960146664, "elapsed_seconds": 16581.30825652799, "env_steps_per_second": 4101.003307337265, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8128092282611762, "eta_seconds": 2926.113221740234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.003800627525015088, "chaser_grad_norm": 0.5278793498873711, "chaser_new_fields": 55.07500076293945, "chaser_partner_vision": 0.9572500586509705, "chaser_return": 16.192501068115234, "chaser_subspace_norm": 222.568115234375, "collisions_per_episode": 14.475000381469727, "entropy": 0.1588267982006073, "explorer_grad_norm": 1.0017587691545486, "explorer_new_fields": 63.95000076293945, "explorer_partner_vision": 0.9572500586509705, "explorer_return": 35.087501525878906, "explorer_subspace_norm": 97.79143524169922, "final_distance": 1.9505478143692017, "max_param_abs": 2.6939685344696045, "policy_loss": -0.0019829598786387237, "value_loss": 3.3157947063446045}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:34:39.568639+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038879.5686545, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.7410687021574835, "elapsed_seconds": 17067.84732776694, "env_steps_per_second": 4101.278776153571, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6772955288796406, "eta_seconds": 2438.263903966706, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0019268124296729638, "chaser_grad_norm": 0.40072858333587646, "chaser_new_fields": 59.04999923706055, "chaser_partner_vision": 0.9667500257492065, "chaser_return": 12.807499885559082, "chaser_subspace_norm": 225.90061950683594, "collisions_per_episode": 10.725000381469727, "entropy": 0.14321690052747726, "explorer_grad_norm": 1.544658899307251, "explorer_new_fields": 67.0999984741211, "explorer_partner_vision": 0.9667500257492065, "explorer_return": 41.16250228881836, "explorer_subspace_norm": 92.53558349609375, "final_distance": 1.992268443107605, "max_param_abs": 2.7126612663269043, "policy_loss": -0.0035469790250033384, "value_loss": 2.651736617088318}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:42:46.107709+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039366.1077204, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.875386900663886, "elapsed_seconds": 17551.392842389992, "env_steps_per_second": 4102.238531525893, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5417096556293207, "eta_seconds": 1950.1547602655546, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.00419491298077368, "chaser_grad_norm": 0.7594298720359802, "chaser_new_fields": 58.22500228881836, "chaser_partner_vision": 0.9647500514984131, "chaser_return": 14.852500915527344, "chaser_subspace_norm": 232.7470245361328, "collisions_per_episode": 12.675000190734863, "entropy": 0.17516248300671577, "explorer_grad_norm": 0.8239959478378296, "explorer_new_fields": 67.0, "explorer_partner_vision": 0.9647500514984131, "explorer_return": 39.962501525878906, "explorer_subspace_norm": 93.31361389160156, "final_distance": 1.801859736442566, "max_param_abs": 2.723428249359131, "policy_loss": -0.003938563286311592, "value_loss": 3.247910737991333}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:50:49.653224+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039849.6532373, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.009863393659439, "elapsed_seconds": 18035.508217173978, "env_steps_per_second": 4103.017176390676, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4062051400264409, "eta_seconds": 1462.3385040951873, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.004114827208930194, "chaser_grad_norm": 0.4523911476135254, "chaser_new_fields": 56.150001525878906, "chaser_partner_vision": 0.9545000195503235, "chaser_return": 14.717500686645508, "chaser_subspace_norm": 214.70062255859375, "collisions_per_episode": 12.925000190734863, "entropy": 0.19482475891709328, "explorer_grad_norm": 1.1241804510354996, "explorer_new_fields": 63.54999923706055, "explorer_partner_vision": 0.9545000195503235, "explorer_return": 35.375, "explorer_subspace_norm": 93.9153823852539, "final_distance": 1.9709968566894531, "max_param_abs": 2.726680040359497, "policy_loss": -0.00388414978522178, "value_loss": 3.2072097659111023}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:58:53.768598+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040333.7686098, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.146812095070556, "elapsed_seconds": 18528.523542254, "env_steps_per_second": 4101.78392394209, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.2708848471089766, "eta_seconds": 975.1854495923158, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0029257981654211562, "chaser_grad_norm": 0.23894650861620903, "chaser_new_fields": 61.97500228881836, "chaser_partner_vision": 0.967750072479248, "chaser_return": 11.65000057220459, "chaser_subspace_norm": 217.26473999023438, "collisions_per_episode": 9.050000190734863, "entropy": 0.14348220080137253, "explorer_grad_norm": 0.7223923951387405, "explorer_new_fields": 69.32500457763672, "explorer_partner_vision": 0.967750072479248, "explorer_return": 46.16250228881836, "explorer_subspace_norm": 95.5700454711914, "final_distance": 1.8984028100967407, "max_param_abs": 2.7633962631225586, "policy_loss": -0.004854720463981721, "value_loss": 2.1610555052757263}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:07:06.783926+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040826.7839417, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.283164775804713, "elapsed_seconds": 19019.39319289697, "env_steps_per_second": 4101.077211502735, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13546576348217215, "eta_seconds": 487.67674853581974, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.005140001692076623, "chaser_grad_norm": 0.701319694519043, "chaser_new_fields": 58.85000228881836, "chaser_partner_vision": 0.9570000171661377, "chaser_return": 11.557500839233398, "chaser_subspace_norm": 257.7256774902344, "collisions_per_episode": 9.574999809265137, "entropy": 0.13568033277988434, "explorer_grad_norm": 0.8242206275463104, "explorer_new_fields": 68.17500305175781, "explorer_partner_vision": 0.9570000171661377, "explorer_return": 43.837501525878906, "explorer_subspace_norm": 94.23152160644531, "final_distance": 1.8717902898788452, "max_param_abs": 2.820730447769165, "policy_loss": -0.0044852363243705895, "value_loss": 2.087482511997223}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:15:17.653576+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041317.6535892, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.416611534084174, "elapsed_seconds": 19499.801522703026, "env_steps_per_second": 4102.605860211368, "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-transformer/social/seed_0008/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:18.061904+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041798.061915, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.416619449098329, "elapsed_seconds": 19499.830016753986, "env_steps_per_second": 4102.599865294472, "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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.003203999427618953, "chaser_grad_norm": 1.5083578824996948, "chaser_new_fields": 57.400001525878906, "chaser_partner_vision": 0.9695000648498535, "chaser_return": 16.552501678466797, "chaser_subspace_norm": 264.9256591796875, "collisions_per_episode": 14.475000381469727, "entropy": 0.1569875180721283, "explorer_grad_norm": 0.9279506653547287, "explorer_new_fields": 64.4000015258789, "explorer_partner_vision": 0.9695000648498535, "explorer_return": 34.45000076293945, "explorer_subspace_norm": 103.04727935791016, "final_distance": 1.7443459033966064, "max_param_abs": 2.8595657348632812, "policy_loss": -0.004643450012451922, "value_loss": 2.9393985867500305}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:18.090398+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041798.0904107, "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..25148acfb22f9a78f03338667367ace8a5f99876 --- /dev/null +++ b/social/seed_0008/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.416619449098329, + "elapsed_seconds": 19499.830016753986, + "env_steps_per_second": 4102.599865294472, + "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-transformer/social/seed_0008/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.003203999427618953, + "chaser_grad_norm": 1.5083578824996948, + "chaser_new_fields": 57.400001525878906, + "chaser_partner_vision": 0.9695000648498535, + "chaser_return": 16.552501678466797, + "chaser_subspace_norm": 264.9256591796875, + "collisions_per_episode": 14.475000381469727, + "entropy": 0.1569875180721283, + "explorer_grad_norm": 0.9279506653547287, + "explorer_new_fields": 64.4000015258789, + "explorer_partner_vision": 0.9695000648498535, + "explorer_return": 34.45000076293945, + "explorer_subspace_norm": 103.04727935791016, + "final_distance": 1.7443459033966064, + "max_param_abs": 2.8595657348632812, + "policy_loss": -0.004643450012451922, + "value_loss": 2.9393985867500305 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:18.090398+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041798.0904107, + "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..1f908a6bb94bcfadc99e4005c6d91ac8ed5f450f --- /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:d50b2b6c17101b2b486d954abdd7be83fe3678f1e69fd0fffeec0607f1fc7685 +size 29760368 diff --git a/social/seed_0009/attempt_01/config.json b/social/seed_0009/attempt_01/config.json new file mode 100644 index 0000000000000000000000000000000000000000..3808a90d011d92c78a8296a8f4bbff65bd9e80f6 --- /dev/null +++ b/social/seed_0009/attempt_01/config.json @@ -0,0 +1,44 @@ +{ + "architecture": "transformer", + "attempt_id": "attempt_01", + "batch_size": 40, + "checkpoint": "runs/mouse-run-run-2-transformer/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-transformer", + "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-transformer/social/seed_0009/attempt_01", + "run_id": "mouse-run-run-2-transformer.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..4dfca8456f9d3348fcb58733a1b5d3bba887ab9e --- /dev/null +++ b/social/seed_0009/attempt_01/metrics.jsonl @@ -0,0 +1,42 @@ +{"device": "cuda", "elapsed_hours": 0.000723327788614875, "elapsed_seconds": 2.6039800390135497, "env_steps_per_second": 1536.1100853581413, "environment_steps": 4000, "error": null, "estimated_cost_usd": null, "eta_hours": 14.465832444508884, "eta_seconds": 52076.99680023198, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.0038027122765242893, "chaser_grad_norm": 5.026018559932709, "chaser_new_fields": 33.60000228881836, "chaser_partner_vision": 0.32500001788139343, "chaser_return": -1.5700002908706665, "chaser_subspace_norm": 0.03227650374174118, "collisions_per_episode": 1.600000023841858, "entropy": 2.75938481092453, "explorer_grad_norm": 6.579230844974518, "explorer_new_fields": 36.17499923706055, "explorer_partner_vision": 0.32500001788139343, "explorer_return": 3.2750000953674316, "explorer_subspace_norm": 0.04388706758618355, "final_distance": 5.103585720062256, "max_param_abs": 1.0012022256851196, "policy_loss": -0.0133217299487427, "value_loss": 7.899519383907318}, "progress_fraction": 5e-05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T19:59:40.989426+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022380.9894402, "update": 1} +{"device": "cuda", "elapsed_hours": 0.1347477191511037, "elapsed_seconds": 485.09178894397337, "env_steps_per_second": 4122.931052603312, "environment_steps": 2000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.255161046893045, "eta_seconds": 18918.57976881496, "healthy": true, "latest_checkpoint": null, "metrics": {"approx_kl": 0.004755442975984181, "chaser_grad_norm": 1.5858053267002106, "chaser_new_fields": 52.650001525878906, "chaser_partner_vision": 0.5205000042915344, "chaser_return": 12.782500267028809, "chaser_subspace_norm": 9.397412300109863, "collisions_per_episode": 11.375, "entropy": 0.6919102221727371, "explorer_grad_norm": 2.1101930737495422, "explorer_new_fields": 62.22500228881836, "explorer_partner_vision": 0.5205000042915344, "explorer_return": 35.28750228881836, "explorer_subspace_norm": 5.533016204833984, "final_distance": 4.548713684082031, "max_param_abs": 1.2132823467254639, "policy_loss": -0.002158123052218741, "value_loss": 3.568888783454895}, "progress_fraction": 0.025, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:07:43.477234+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783022863.4772465, "update": 500} +{"device": "cuda", "elapsed_hours": 0.2689785306333215, "elapsed_seconds": 968.3227102799574, "env_steps_per_second": 4130.854267420349, "environment_steps": 4000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.110592082033109, "eta_seconds": 18398.13149531919, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.0076650890790865844, "chaser_grad_norm": 1.4071732312440872, "chaser_new_fields": 55.60000228881836, "chaser_partner_vision": 0.5492500066757202, "chaser_return": 11.637499809265137, "chaser_subspace_norm": 16.199703216552734, "collisions_per_episode": 9.875, "entropy": 0.5089846551418304, "explorer_grad_norm": 1.4974915981292725, "explorer_new_fields": 64.42500305175781, "explorer_partner_vision": 0.5492500066757202, "explorer_return": 38.92499923706055, "explorer_subspace_norm": 10.612075805664062, "final_distance": 4.761884689331055, "max_param_abs": 1.3461778163909912, "policy_loss": -0.00339932309324098, "value_loss": 3.4826159477233887}, "progress_fraction": 0.05, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:15:46.708154+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023346.7081683, "update": 1000} +{"device": "cuda", "elapsed_hours": 0.40554712050193403, "elapsed_seconds": 1459.9696338069625, "env_steps_per_second": 4109.6745172395285, "environment_steps": 6000000, "error": null, "estimated_cost_usd": null, "eta_hours": 5.001747819523853, "eta_seconds": 18006.29215028587, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_001000.safetensors", "metrics": {"approx_kl": 0.00543786848936012, "chaser_grad_norm": 3.2535484433174133, "chaser_new_fields": 54.275001525878906, "chaser_partner_vision": 0.5082499980926514, "chaser_return": 10.067500114440918, "chaser_subspace_norm": 20.757003784179688, "collisions_per_episode": 8.625, "entropy": 0.47901443392038345, "explorer_grad_norm": 2.215997964143753, "explorer_new_fields": 69.625, "explorer_partner_vision": 0.5082499980926514, "explorer_return": 47.462501525878906, "explorer_subspace_norm": 15.635637283325195, "final_distance": 4.485320568084717, "max_param_abs": 1.4825319051742554, "policy_loss": -0.005668422083360269, "value_loss": 3.3295662999153137}, "progress_fraction": 0.075, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:23:58.355080+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783023838.3550937, "update": 1500} +{"device": "cuda", "elapsed_hours": 0.5423076709408391, "elapsed_seconds": 1952.3076153870206, "env_steps_per_second": 4097.714897462048, "environment_steps": 8000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.880769038467552, "eta_seconds": 17570.768538483186, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.002811812950870518, "chaser_grad_norm": 2.022128641605377, "chaser_new_fields": 50.42499923706055, "chaser_partner_vision": 0.6205000281333923, "chaser_return": 13.375, "chaser_subspace_norm": 27.18659210205078, "collisions_per_episode": 12.449999809265137, "entropy": 0.4175580069422722, "explorer_grad_norm": 1.7573799788951874, "explorer_new_fields": 68.42500305175781, "explorer_partner_vision": 0.6205000281333923, "explorer_return": 42.92499923706055, "explorer_subspace_norm": 21.36162567138672, "final_distance": 2.8846395015716553, "max_param_abs": 1.5631178617477417, "policy_loss": -0.00794941140897265, "value_loss": 3.686682879924774}, "progress_fraction": 0.1, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:32:10.693061+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024330.693076, "update": 2000} +{"device": "cuda", "elapsed_hours": 0.6783107926861137, "elapsed_seconds": 2441.9188536700094, "env_steps_per_second": 4095.1401742817116, "environment_steps": 10000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.748175548802796, "eta_seconds": 17093.431975690066, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_002000.safetensors", "metrics": {"approx_kl": 0.006651435111915793, "chaser_grad_norm": 1.111131489276886, "chaser_new_fields": 48.72500228881836, "chaser_partner_vision": 0.8250000476837158, "chaser_return": 17.93000030517578, "chaser_subspace_norm": 31.41754722595215, "collisions_per_episode": 17.200000762939453, "entropy": 0.3929666131734848, "explorer_grad_norm": 1.1966542303562164, "explorer_new_fields": 64.25, "explorer_partner_vision": 0.8250000476837158, "explorer_return": 32.67499923706055, "explorer_subspace_norm": 24.74244499206543, "final_distance": 1.982876181602478, "max_param_abs": 1.6375449895858765, "policy_loss": -0.008235584467641388, "value_loss": 3.632794439792633}, "progress_fraction": 0.125, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:40:20.304298+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783024820.3043084, "update": 2500} +{"device": "cuda", "elapsed_hours": 0.8154701736927705, "elapsed_seconds": 2935.692625293974, "env_steps_per_second": 4087.62140034955, "environment_steps": 12000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.620997650925699, "eta_seconds": 16635.591543332517, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.006007212468618017, "chaser_grad_norm": 1.0740922093391418, "chaser_new_fields": 48.625, "chaser_partner_vision": 0.9240000247955322, "chaser_return": 17.242502212524414, "chaser_subspace_norm": 46.90514373779297, "collisions_per_episode": 16.625, "entropy": 0.30619022250175476, "explorer_grad_norm": 1.9078352749347687, "explorer_new_fields": 59.07500076293945, "explorer_partner_vision": 0.9240000247955322, "explorer_return": 24.337499618530273, "explorer_subspace_norm": 30.008319854736328, "final_distance": 1.727256417274475, "max_param_abs": 1.6937978267669678, "policy_loss": -0.007849829389712681, "value_loss": 3.5645273327827454}, "progress_fraction": 0.15, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:48:34.078071+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025314.0780857, "update": 3000} +{"device": "cuda", "elapsed_hours": 0.9493466281286156, "elapsed_seconds": 3417.6478612630162, "env_steps_per_second": 4096.384580366393, "environment_steps": 14000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.475491246892045, "eta_seconds": 16111.768488811362, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_003000.safetensors", "metrics": {"approx_kl": 0.006429051130812269, "chaser_grad_norm": 0.8913671672344208, "chaser_new_fields": 51.04999923706055, "chaser_partner_vision": 0.924750030040741, "chaser_return": 21.747499465942383, "chaser_subspace_norm": 51.39407730102539, "collisions_per_episode": 20.17500114440918, "entropy": 0.3543940633535385, "explorer_grad_norm": 1.9905150830745697, "explorer_new_fields": 59.125, "explorer_partner_vision": 0.924750030040741, "explorer_return": 23.537500381469727, "explorer_subspace_norm": 30.78665542602539, "final_distance": 2.295875310897827, "max_param_abs": 1.7198820114135742, "policy_loss": -0.007569677537318853, "value_loss": 4.241915345191956}, "progress_fraction": 0.175, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T20:56:36.033305+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783025796.0333173, "update": 3500} +{"device": "cuda", "elapsed_hours": 1.0862288366124913, "elapsed_seconds": 3910.423811804969, "env_steps_per_second": 4091.628112456368, "environment_steps": 16000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.344915346449965, "eta_seconds": 15641.695247219875, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.008145414923767191, "chaser_grad_norm": 0.7966946363449097, "chaser_new_fields": 52.525001525878906, "chaser_partner_vision": 0.9382500648498535, "chaser_return": 18.91750144958496, "chaser_subspace_norm": 62.55392074584961, "collisions_per_episode": 17.524999618530273, "entropy": 0.30151088535785675, "explorer_grad_norm": 1.524125725030899, "explorer_new_fields": 60.775001525878906, "explorer_partner_vision": 0.9382500648498535, "explorer_return": 27.075000762939453, "explorer_subspace_norm": 30.9876708984375, "final_distance": 2.4827067852020264, "max_param_abs": 1.7440770864486694, "policy_loss": -0.00923282765222666, "value_loss": 3.694827914237976}, "progress_fraction": 0.2, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:04:48.809256+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026288.8092704, "update": 4000} +{"device": "cuda", "elapsed_hours": 1.2223848155410895, "elapsed_seconds": 4400.585335947922, "env_steps_per_second": 4090.364945990225, "environment_steps": 18000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.210436586863753, "eta_seconds": 15157.57171270951, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_004000.safetensors", "metrics": {"approx_kl": 0.004178010998771242, "chaser_grad_norm": 0.7092046290636063, "chaser_new_fields": 54.04999923706055, "chaser_partner_vision": 0.9587500691413879, "chaser_return": 18.297500610351562, "chaser_subspace_norm": 68.53948211669922, "collisions_per_episode": 16.475000381469727, "entropy": 0.26812686771154404, "explorer_grad_norm": 1.3655212074518204, "explorer_new_fields": 59.70000076293945, "explorer_partner_vision": 0.9587500691413879, "explorer_return": 26.0625, "explorer_subspace_norm": 37.85118865966797, "final_distance": 1.6317100524902344, "max_param_abs": 1.7602100372314453, "policy_loss": -0.004626329940748519, "value_loss": 3.492742896080017}, "progress_fraction": 0.225, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:12:58.970781+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783026778.9707952, "update": 4500} +{"device": "cuda", "elapsed_hours": 1.359833971754706, "elapsed_seconds": 4895.402298316942, "env_steps_per_second": 4085.466072293196, "environment_steps": 20000000, "error": null, "estimated_cost_usd": null, "eta_hours": 4.079501915264118, "eta_seconds": 14686.206894950825, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.00920238255236594, "chaser_grad_norm": 0.7991169765591621, "chaser_new_fields": 53.775001525878906, "chaser_partner_vision": 0.9600000381469727, "chaser_return": 19.510000228881836, "chaser_subspace_norm": 78.4761962890625, "collisions_per_episode": 17.850000381469727, "entropy": 0.20775993913412094, "explorer_grad_norm": 1.4601047039031982, "explorer_new_fields": 59.75, "explorer_partner_vision": 0.9600000381469727, "explorer_return": 24.850000381469727, "explorer_subspace_norm": 45.24058532714844, "final_distance": 1.8044236898422241, "max_param_abs": 1.792447805404663, "policy_loss": -0.00485417765285856, "value_loss": 3.738081157207489}, "progress_fraction": 0.25, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:21:13.787745+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027273.7877605, "update": 5000} +{"device": "cuda", "elapsed_hours": 1.4940944158780622, "elapsed_seconds": 5378.739897161024, "env_steps_per_second": 4090.1773316110557, "environment_steps": 22000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.9389761873148914, "eta_seconds": 14180.314274333608, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_005000.safetensors", "metrics": {"approx_kl": 0.005502791298802379, "chaser_grad_norm": 0.4452984258532524, "chaser_new_fields": 53.75, "chaser_partner_vision": 0.9610000252723694, "chaser_return": 19.44499969482422, "chaser_subspace_norm": 87.14356994628906, "collisions_per_episode": 17.600000381469727, "entropy": 0.2977784052491188, "explorer_grad_norm": 1.4750292003154755, "explorer_new_fields": 62.900001525878906, "explorer_partner_vision": 0.9610000252723694, "explorer_return": 29.850000381469727, "explorer_subspace_norm": 37.74265670776367, "final_distance": 1.6021215915679932, "max_param_abs": 1.8037797212600708, "policy_loss": -0.004444488349832909, "value_loss": 3.653634786605835}, "progress_fraction": 0.275, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:29:17.125341+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783027757.125352, "update": 5500} +{"device": "cuda", "elapsed_hours": 1.6297460299658189, "elapsed_seconds": 5867.085707876948, "env_steps_per_second": 4090.6169084556623, "environment_steps": 24000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.80274073658691, "eta_seconds": 13689.866651712877, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.003138580170542582, "chaser_grad_norm": 0.4320807158946991, "chaser_new_fields": 56.04999923706055, "chaser_partner_vision": 0.9690000414848328, "chaser_return": 16.267499923706055, "chaser_subspace_norm": 109.2682113647461, "collisions_per_episode": 14.375, "entropy": 0.16355393454432487, "explorer_grad_norm": 1.1077640056610107, "explorer_new_fields": 66.625, "explorer_partner_vision": 0.9690000414848328, "explorer_return": 36.150001525878906, "explorer_subspace_norm": 45.20527267456055, "final_distance": 1.470686912536621, "max_param_abs": 1.8184922933578491, "policy_loss": -0.005340472675861463, "value_loss": 2.813709020614624}, "progress_fraction": 0.3, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:37:25.471153+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028245.471169, "update": 6000} +{"device": "cuda", "elapsed_hours": 1.7643791413316567, "elapsed_seconds": 6351.764908793964, "env_steps_per_second": 4093.3504897203015, "environment_steps": 26000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.6644797550734407, "eta_seconds": 13192.127118264387, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_006000.safetensors", "metrics": {"approx_kl": 0.005023273031266484, "chaser_grad_norm": 0.39883723855018616, "chaser_new_fields": 52.75, "chaser_partner_vision": 0.9625000357627869, "chaser_return": 14.267500877380371, "chaser_subspace_norm": 104.88886260986328, "collisions_per_episode": 12.975000381469727, "entropy": 0.2776241600513458, "explorer_grad_norm": 0.9978158175945282, "explorer_new_fields": 60.79999923706055, "explorer_partner_vision": 0.9625000357627869, "explorer_return": 31.5625, "explorer_subspace_norm": 36.82559585571289, "final_distance": 1.635999321937561, "max_param_abs": 1.825235366821289, "policy_loss": -0.007816974651078112, "value_loss": 3.680994212627411}, "progress_fraction": 0.325, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:45:30.150354+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783028730.1503677, "update": 6500} +{"device": "cuda", "elapsed_hours": 1.90069195498334, "elapsed_seconds": 6842.491037940024, "env_steps_per_second": 4092.0769709081824, "environment_steps": 28000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.529856487826203, "eta_seconds": 12707.48335617433, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.004577269709015397, "chaser_grad_norm": 0.42832470685243607, "chaser_new_fields": 55.20000076293945, "chaser_partner_vision": 0.9630000591278076, "chaser_return": 15.114999771118164, "chaser_subspace_norm": 126.4784164428711, "collisions_per_episode": 13.40000057220459, "entropy": 0.20087017118930817, "explorer_grad_norm": 1.489731252193451, "explorer_new_fields": 62.150001525878906, "explorer_partner_vision": 0.9630000591278076, "explorer_return": 32.10000228881836, "explorer_subspace_norm": 45.85857009887695, "final_distance": 1.7384480237960815, "max_param_abs": 1.8486905097961426, "policy_loss": -0.006352735267029175, "value_loss": 3.0404184460639954}, "progress_fraction": 0.35, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T21:53:40.876483+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029220.8764968, "update": 7000} +{"device": "cuda", "elapsed_hours": 2.040538195431646, "elapsed_seconds": 7345.937503553927, "env_steps_per_second": 4083.8899031588758, "environment_steps": 30000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.4008969923860772, "eta_seconds": 12243.229172589878, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_007000.safetensors", "metrics": {"approx_kl": 0.005035266689335405, "chaser_grad_norm": 0.6151941567659378, "chaser_new_fields": 45.25, "chaser_partner_vision": 0.9500000476837158, "chaser_return": 20.140003204345703, "chaser_subspace_norm": 126.91571807861328, "collisions_per_episode": 19.649999618530273, "entropy": 0.23392390832304955, "explorer_grad_norm": 0.7963254600763321, "explorer_new_fields": 51.45000076293945, "explorer_partner_vision": 0.9500000476837158, "explorer_return": 14.612500190734863, "explorer_subspace_norm": 49.23459243774414, "final_distance": 1.7384439706802368, "max_param_abs": 1.8811355829238892, "policy_loss": -0.00343825779761886, "value_loss": 4.618174314498901}, "progress_fraction": 0.375, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:02:04.322947+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783029724.322959, "update": 7500} +{"device": "cuda", "elapsed_hours": 2.180488174737774, "elapsed_seconds": 7849.757429055986, "env_steps_per_second": 4076.55908978175, "environment_steps": 32000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.270732262106661, "eta_seconds": 11774.636143583979, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.0020045672982913487, "chaser_grad_norm": 1.9626288414001465, "chaser_new_fields": 52.17499923706055, "chaser_partner_vision": 0.968000054359436, "chaser_return": 18.957502365112305, "chaser_subspace_norm": 128.3649139404297, "collisions_per_episode": 17.575000762939453, "entropy": 0.23982753977179527, "explorer_grad_norm": 0.7570929825305939, "explorer_new_fields": 58.125, "explorer_partner_vision": 0.968000054359436, "explorer_return": 23.1875, "explorer_subspace_norm": 46.636356353759766, "final_distance": 1.768707275390625, "max_param_abs": 1.9123574495315552, "policy_loss": -0.005071938496101325, "value_loss": 3.8649072647094727}, "progress_fraction": 0.4, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:10:28.142874+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030228.1428878, "update": 8000} +{"device": "cuda", "elapsed_hours": 2.3155448580758335, "elapsed_seconds": 8335.961489073, "env_steps_per_second": 4078.7136606338813, "environment_steps": 34000000, "error": null, "estimated_cost_usd": null, "eta_hours": 3.1327959844555395, "eta_seconds": 11278.065544039942, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_008000.safetensors", "metrics": {"approx_kl": 0.004160874696051309, "chaser_grad_norm": 0.7958575785160065, "chaser_new_fields": 53.375, "chaser_partner_vision": 0.9630000591278076, "chaser_return": 18.24250030517578, "chaser_subspace_norm": 138.3916015625, "collisions_per_episode": 16.625, "entropy": 0.16832594200968742, "explorer_grad_norm": 1.1085042357444763, "explorer_new_fields": 61.47500228881836, "explorer_partner_vision": 0.9630000591278076, "explorer_return": 28.350000381469727, "explorer_subspace_norm": 54.24068832397461, "final_distance": 1.605097770690918, "max_param_abs": 1.97091543674469, "policy_loss": -0.003103218309579603, "value_loss": 2.933476507663727}, "progress_fraction": 0.425, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:18:34.346933+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783030714.3469443, "update": 8500} +{"device": "cuda", "elapsed_hours": 2.44895817809528, "elapsed_seconds": 8816.249441143009, "env_steps_per_second": 4083.369038085278, "environment_steps": 36000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.993171106560898, "eta_seconds": 10775.415983619234, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.003159792652279947, "chaser_grad_norm": 0.5041773095726967, "chaser_new_fields": 54.32500076293945, "chaser_partner_vision": 0.9697500467300415, "chaser_return": 16.1200008392334, "chaser_subspace_norm": 158.76690673828125, "collisions_per_episode": 14.550000190734863, "entropy": 0.18282336741685867, "explorer_grad_norm": 0.8686060458421707, "explorer_new_fields": 64.5999984741211, "explorer_partner_vision": 0.9697500467300415, "explorer_return": 33.13750076293945, "explorer_subspace_norm": 55.67268753051758, "final_distance": 1.6979671716690063, "max_param_abs": 1.9921530485153198, "policy_loss": -0.00536713201273642, "value_loss": 2.8779627084732056}, "progress_fraction": 0.45, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:26:34.634887+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031194.634901, "update": 9000} +{"device": "cuda", "elapsed_hours": 2.583117585490561, "elapsed_seconds": 9299.22330776602, "env_steps_per_second": 4086.3627791651397, "environment_steps": 38000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.8550246997527258, "eta_seconds": 10278.088919109812, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_009000.safetensors", "metrics": {"approx_kl": 0.00420251796749227, "chaser_grad_norm": 0.8840731233358383, "chaser_new_fields": 53.67499923706055, "chaser_partner_vision": 0.968500018119812, "chaser_return": 13.635001182556152, "chaser_subspace_norm": 157.84275817871094, "collisions_per_episode": 12.350000381469727, "entropy": 0.18863104656338692, "explorer_grad_norm": 1.4644344747066498, "explorer_new_fields": 59.92499923706055, "explorer_partner_vision": 0.968500018119812, "explorer_return": 29.587499618530273, "explorer_subspace_norm": 59.349334716796875, "final_distance": 1.5879405736923218, "max_param_abs": 2.0039682388305664, "policy_loss": -0.004914777579713814, "value_loss": 3.4356402158737183}, "progress_fraction": 0.475, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:34:37.608751+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783031677.608764, "update": 9500} +{"device": "cuda", "elapsed_hours": 2.716491514392773, "elapsed_seconds": 9779.369451813982, "env_steps_per_second": 4090.2432612953767, "environment_steps": 40000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.716491514392773, "eta_seconds": 9779.369451813982, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.0040675974684156735, "chaser_grad_norm": 0.8714076429605484, "chaser_new_fields": 54.775001525878906, "chaser_partner_vision": 0.9815000295639038, "chaser_return": 13.927499771118164, "chaser_subspace_norm": 158.25289916992188, "collisions_per_episode": 12.425000190734863, "entropy": 0.1915062628686428, "explorer_grad_norm": 1.0908084958791733, "explorer_new_fields": 61.25, "explorer_partner_vision": 0.9815000295639038, "explorer_return": 30.86250114440918, "explorer_subspace_norm": 60.13104248046875, "final_distance": 1.6341546773910522, "max_param_abs": 2.016425609588623, "policy_loss": -0.006213883706365664, "value_loss": 2.7247623801231384}, "progress_fraction": 0.5, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:42:37.754898+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032157.7549126, "update": 10000} +{"device": "cuda", "elapsed_hours": 2.8516560393308303, "elapsed_seconds": 10265.961741590989, "env_steps_per_second": 4091.1899982875802, "environment_steps": 42000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.580069749870751, "eta_seconds": 9288.251099534704, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_010000.safetensors", "metrics": {"approx_kl": 0.005137112095198404, "chaser_grad_norm": 0.41913997381925583, "chaser_new_fields": 50.60000228881836, "chaser_partner_vision": 0.9567500352859497, "chaser_return": 14.15000057220459, "chaser_subspace_norm": 165.38671875, "collisions_per_episode": 13.40000057220459, "entropy": 0.18020926788449287, "explorer_grad_norm": 0.7766301780939102, "explorer_new_fields": 59.32500076293945, "explorer_partner_vision": 0.9567500352859497, "explorer_return": 28.462499618530273, "explorer_subspace_norm": 62.46317672729492, "final_distance": 1.585931420326233, "max_param_abs": 2.047652006149292, "policy_loss": -0.006081848105409904, "value_loss": 3.6701528429985046}, "progress_fraction": 0.525, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:50:44.347185+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783032644.3471975, "update": 10500} +{"device": "cuda", "elapsed_hours": 2.9871363096305545, "elapsed_seconds": 10753.690714669996, "env_steps_per_second": 4091.618511956641, "environment_steps": 44000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.4440206169704535, "eta_seconds": 8798.474221093633, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004140301480955344, "chaser_grad_norm": 0.5378393009305, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.9670000672340393, "chaser_return": 13.089999198913574, "chaser_subspace_norm": 183.41555786132812, "collisions_per_episode": 11.100000381469727, "entropy": 0.1406456083059311, "explorer_grad_norm": 1.318773239850998, "explorer_new_fields": 64.80000305175781, "explorer_partner_vision": 0.9670000672340393, "explorer_return": 38.72500228881836, "explorer_subspace_norm": 70.11807250976562, "final_distance": 1.6715869903564453, "max_param_abs": 2.131923198699951, "policy_loss": -0.005167009411394474, "value_loss": 2.9128845930099487}, "progress_fraction": 0.55, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T22:58:52.076160+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033132.0761752, "update": 11000} +{"device": "cuda", "elapsed_hours": 3.1212548896069507, "elapsed_seconds": 11236.517602585023, "env_steps_per_second": 4093.795037478288, "environment_steps": 46000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.3070144836225293, "eta_seconds": 8305.252141041105, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_011000.safetensors", "metrics": {"approx_kl": 0.004444854070307258, "chaser_grad_norm": 0.4114886522293091, "chaser_new_fields": 56.0, "chaser_partner_vision": 0.9495000243186951, "chaser_return": 12.05250072479248, "chaser_subspace_norm": 167.727783203125, "collisions_per_episode": 10.324999809265137, "entropy": 0.21219957992434502, "explorer_grad_norm": 1.2058109641075134, "explorer_new_fields": 61.35000228881836, "explorer_partner_vision": 0.9495000243186951, "explorer_return": 33.9375, "explorer_subspace_norm": 67.38232421875, "final_distance": 1.8433490991592407, "max_param_abs": 2.2799575328826904, "policy_loss": -0.0053471988085789235, "value_loss": 3.392748177051544}, "progress_fraction": 0.575, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:06:54.903047+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783033614.9030585, "update": 11500} +{"device": "cuda", "elapsed_hours": 3.2587898411847225, "elapsed_seconds": 11731.643428265, "env_steps_per_second": 4091.4983730543495, "environment_steps": 48000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.172526560789815, "eta_seconds": 7821.095618843334, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.002780596478253905, "chaser_grad_norm": 1.0835029482841492, "chaser_new_fields": 57.82500076293945, "chaser_partner_vision": 0.9412500262260437, "chaser_return": 9.127500534057617, "chaser_subspace_norm": 180.20083618164062, "collisions_per_episode": 7.275000095367432, "entropy": 0.1915491260588169, "explorer_grad_norm": 1.7708372473716736, "explorer_new_fields": 66.30000305175781, "explorer_partner_vision": 0.9412500262260437, "explorer_return": 43.07500076293945, "explorer_subspace_norm": 62.17178726196289, "final_distance": 1.76237952709198, "max_param_abs": 2.2569639682769775, "policy_loss": -0.005818219148551407, "value_loss": 2.7338308095932007}, "progress_fraction": 0.6, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:15:10.028873+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034110.028888, "update": 12000} +{"device": "cuda", "elapsed_hours": 3.394498959807485, "elapsed_seconds": 12220.196255306946, "env_steps_per_second": 4091.5873162254793, "environment_steps": 50000000, "error": null, "estimated_cost_usd": null, "eta_hours": 2.036699375884491, "eta_seconds": 7332.117753184168, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_012000.safetensors", "metrics": {"approx_kl": 0.009038661571096895, "chaser_grad_norm": 0.5724573209881783, "chaser_new_fields": 52.92499923706055, "chaser_partner_vision": 0.968000054359436, "chaser_return": 17.85500144958496, "chaser_subspace_norm": 165.585693359375, "collisions_per_episode": 16.399999618530273, "entropy": 0.23524165898561478, "explorer_grad_norm": 1.2622281908988953, "explorer_new_fields": 60.10000228881836, "explorer_partner_vision": 0.968000054359436, "explorer_return": 26.92500114440918, "explorer_subspace_norm": 64.0269546508789, "final_distance": 2.2381088733673096, "max_param_abs": 2.300077199935913, "policy_loss": -0.006312312293857758, "value_loss": 3.993464231491089}, "progress_fraction": 0.625, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:23:18.581698+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783034598.5817106, "update": 12500} +{"device": "cuda", "elapsed_hours": 3.5284770946374757, "elapsed_seconds": 12702.517540694913, "env_steps_per_second": 4093.676693097111, "environment_steps": 52000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.8999492048047948, "eta_seconds": 6839.817137297261, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.0027599246014631262, "chaser_grad_norm": 0.7339675948023796, "chaser_new_fields": 59.20000076293945, "chaser_partner_vision": 0.9582500457763672, "chaser_return": 9.952499389648438, "chaser_subspace_norm": 167.568359375, "collisions_per_episode": 7.875, "entropy": 0.2076490931212902, "explorer_grad_norm": 0.8425474464893341, "explorer_new_fields": 67.4749984741211, "explorer_partner_vision": 0.9582500457763672, "explorer_return": 43.9375, "explorer_subspace_norm": 57.674495697021484, "final_distance": 1.6045942306518555, "max_param_abs": 2.307316780090332, "policy_loss": -0.005739407592457724, "value_loss": 2.9169874787330627}, "progress_fraction": 0.65, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:31:20.902987+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035080.9030015, "update": 13000} +{"device": "cuda", "elapsed_hours": 3.6647399410263946, "elapsed_seconds": 13193.06378769502, "env_steps_per_second": 4093.0598736561114, "environment_steps": 54000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.7645044160497456, "eta_seconds": 6352.215897779084, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_013000.safetensors", "metrics": {"approx_kl": 0.003413488739742121, "chaser_grad_norm": 0.5240671336650848, "chaser_new_fields": 58.025001525878906, "chaser_partner_vision": 0.9492500424385071, "chaser_return": 10.36500072479248, "chaser_subspace_norm": 188.77194213867188, "collisions_per_episode": 8.40000057220459, "entropy": 0.20054690167307854, "explorer_grad_norm": 1.2580983936786652, "explorer_new_fields": 66.4000015258789, "explorer_partner_vision": 0.9492500424385071, "explorer_return": 42.25, "explorer_subspace_norm": 62.18594741821289, "final_distance": 1.8519560098648071, "max_param_abs": 2.3287360668182373, "policy_loss": -0.0049286139579436394, "value_loss": 2.942625105381012}, "progress_fraction": 0.675, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:39:31.449233+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783035571.4492455, "update": 13500} +{"device": "cuda", "elapsed_hours": 3.798413469242482, "elapsed_seconds": 13674.288489272934, "env_steps_per_second": 4095.2770627832156, "environment_steps": 56000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.6278914868182066, "eta_seconds": 5860.409352545544, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.0031029039100363143, "chaser_grad_norm": 0.524135485291481, "chaser_new_fields": 55.32500076293945, "chaser_partner_vision": 0.9585000276565552, "chaser_return": 12.02500057220459, "chaser_subspace_norm": 207.37091064453125, "collisions_per_episode": 10.550000190734863, "entropy": 0.16616593673825264, "explorer_grad_norm": 0.7165043205022812, "explorer_new_fields": 66.4749984741211, "explorer_partner_vision": 0.9585000276565552, "explorer_return": 40.3125, "explorer_subspace_norm": 76.35435485839844, "final_distance": 1.5791975259780884, "max_param_abs": 2.3233821392059326, "policy_loss": -0.0064672464122850215, "value_loss": 2.9724830389022827}, "progress_fraction": 0.7, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:47:32.673936+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036052.673952, "update": 14000} +{"device": "cuda", "elapsed_hours": 3.9349341308261176, "elapsed_seconds": 14165.762870974024, "env_steps_per_second": 4094.378857551212, "environment_steps": 58000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.4925612220374929, "eta_seconds": 5373.220399334974, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_014000.safetensors", "metrics": {"approx_kl": 0.001890498524655726, "chaser_grad_norm": 0.43047017604112625, "chaser_new_fields": 57.22500228881836, "chaser_partner_vision": 0.9692500233650208, "chaser_return": 11.59000015258789, "chaser_subspace_norm": 186.98597717285156, "collisions_per_episode": 9.90000057220459, "entropy": 0.1543101966381073, "explorer_grad_norm": 0.8024827539920807, "explorer_new_fields": 64.07500457763672, "explorer_partner_vision": 0.9692500233650208, "explorer_return": 37.60000228881836, "explorer_subspace_norm": 78.98009490966797, "final_distance": 1.997841715812683, "max_param_abs": 2.344755172729492, "policy_loss": -0.006300987407556136, "value_loss": 2.550761580467224}, "progress_fraction": 0.725, "schema_version": 2, "state": "running", "timestamp": "2026-07-02T23:55:44.148315+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783036544.1483278, "update": 14500} +{"device": "cuda", "elapsed_hours": 4.06966967909917, "elapsed_seconds": 14650.810844757012, "env_steps_per_second": 4095.3364722111473, "environment_steps": 60000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.3565565596997233, "eta_seconds": 4883.603614919004, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0067666837250524026, "chaser_grad_norm": 0.4584822431206703, "chaser_new_fields": 59.900001525878906, "chaser_partner_vision": 0.9652500748634338, "chaser_return": 10.822500228881836, "chaser_subspace_norm": 204.42803955078125, "collisions_per_episode": 8.52500057220459, "entropy": 0.16078611463308334, "explorer_grad_norm": 0.7388524413108826, "explorer_new_fields": 68.5999984741211, "explorer_partner_vision": 0.9652500748634338, "explorer_return": 45.67499923706055, "explorer_subspace_norm": 73.9074478149414, "final_distance": 2.005887508392334, "max_param_abs": 2.369492530822754, "policy_loss": -0.005749120077098269, "value_loss": 2.352930247783661}, "progress_fraction": 0.75, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:03:49.196290+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037029.1963043, "update": 15000} +{"device": "cuda", "elapsed_hours": 4.20916648086694, "elapsed_seconds": 15152.999331120984, "env_steps_per_second": 4091.5992039058174, "environment_steps": 62000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.2220160750904019, "eta_seconds": 4399.257870325447, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_015000.safetensors", "metrics": {"approx_kl": 0.0036926642350927352, "chaser_grad_norm": 0.5912124216556549, "chaser_new_fields": 59.70000076293945, "chaser_partner_vision": 0.9537500739097595, "chaser_return": 11.547500610351562, "chaser_subspace_norm": 228.60569763183594, "collisions_per_episode": 9.324999809265137, "entropy": 0.14043940231204033, "explorer_grad_norm": 0.8066128939390182, "explorer_new_fields": 68.95000457763672, "explorer_partner_vision": 0.9537500739097595, "explorer_return": 44.900001525878906, "explorer_subspace_norm": 77.73728942871094, "final_distance": 1.7932811975479126, "max_param_abs": 2.3865718841552734, "policy_loss": -0.004267201186166858, "value_loss": 2.7086270451545715}, "progress_fraction": 0.775, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:12:11.384775+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783037531.3847864, "update": 15500} +{"device": "cuda", "elapsed_hours": 4.346961487734451, "elapsed_seconds": 15649.061355844024, "env_steps_per_second": 4089.7021581489093, "environment_steps": 64000000, "error": null, "estimated_cost_usd": null, "eta_hours": 1.0867403719336128, "eta_seconds": 3912.265338961006, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.00899587140769853, "chaser_grad_norm": 1.8401049375534058, "chaser_new_fields": 51.25, "chaser_partner_vision": 0.9582500457763672, "chaser_return": 20.71750259399414, "chaser_subspace_norm": 213.59378051757812, "collisions_per_episode": 19.225000381469727, "entropy": 0.13754115253686905, "explorer_grad_norm": 1.3938065767288208, "explorer_new_fields": 57.375, "explorer_partner_vision": 0.9582500457763672, "explorer_return": 21.875, "explorer_subspace_norm": 85.4321060180664, "final_distance": 1.6317100524902344, "max_param_abs": 2.3802859783172607, "policy_loss": 0.0005846844140933172, "value_loss": 4.14808976650238}, "progress_fraction": 0.8, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:20:27.446802+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038027.446817, "update": 16000} +{"device": "cuda", "elapsed_hours": 4.484853996562209, "elapsed_seconds": 16145.474387623952, "env_steps_per_second": 4087.832814041761, "environment_steps": 66000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.9513326659374381, "eta_seconds": 3424.7975973747775, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_016000.safetensors", "metrics": {"approx_kl": 0.0018162342323422287, "chaser_grad_norm": 0.25812123715877533, "chaser_new_fields": 58.29999923706055, "chaser_partner_vision": 0.9785000681877136, "chaser_return": 12.367500305175781, "chaser_subspace_norm": 213.47239685058594, "collisions_per_episode": 10.375, "entropy": 0.1525474339723587, "explorer_grad_norm": 1.0173778980970383, "explorer_new_fields": 67.1500015258789, "explorer_partner_vision": 0.9785000681877136, "explorer_return": 41.60000228881836, "explorer_subspace_norm": 85.16871643066406, "final_distance": 1.702092170715332, "max_param_abs": 2.3969850540161133, "policy_loss": -0.0048129543024231225, "value_loss": 2.4122917652130127}, "progress_fraction": 0.825, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:28:43.859831+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783038523.8598433, "update": 16500} +{"device": "cuda", "elapsed_hours": 4.621616250113035, "elapsed_seconds": 16637.818500406924, "env_steps_per_second": 4087.074275893614, "environment_steps": 68000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.8155793382552413, "eta_seconds": 2936.085617718869, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.0034244980526558777, "chaser_grad_norm": 0.6200611591339111, "chaser_new_fields": 59.150001525878906, "chaser_partner_vision": 0.9585000276565552, "chaser_return": 10.800000190734863, "chaser_subspace_norm": 236.82606506347656, "collisions_per_episode": 8.699999809265137, "entropy": 0.13780700415372849, "explorer_grad_norm": 0.7542365491390228, "explorer_new_fields": 68.2750015258789, "explorer_partner_vision": 0.9585000276565552, "explorer_return": 44.875, "explorer_subspace_norm": 90.64749908447266, "final_distance": 2.0387020111083984, "max_param_abs": 2.466871976852417, "policy_loss": -0.004792424348188717, "value_loss": 2.620694637298584}, "progress_fraction": 0.85, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:36:56.203946+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039016.2039595, "update": 17000} +{"device": "cuda", "elapsed_hours": 4.761001563909974, "elapsed_seconds": 17139.60563007591, "env_steps_per_second": 4084.107972540905, "environment_steps": 70000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.6801430805585679, "eta_seconds": 2448.5150900108442, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_017000.safetensors", "metrics": {"approx_kl": 0.003942118872373612, "chaser_grad_norm": 0.7372853010892868, "chaser_new_fields": 55.775001525878906, "chaser_partner_vision": 0.9622500538825989, "chaser_return": 15.642500877380371, "chaser_subspace_norm": 240.84996032714844, "collisions_per_episode": 13.824999809265137, "entropy": 0.13363275304436684, "explorer_grad_norm": 0.621310368180275, "explorer_new_fields": 64.57500457763672, "explorer_partner_vision": 0.9622500538825989, "explorer_return": 35.337501525878906, "explorer_subspace_norm": 90.43130493164062, "final_distance": 1.957438349723816, "max_param_abs": 2.4603312015533447, "policy_loss": -0.00399300978891759, "value_loss": 3.2278329133987427}, "progress_fraction": 0.875, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:45:17.991074+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783039517.9910872, "update": 17500} +{"device": "cuda", "elapsed_hours": 4.896728742249445, "elapsed_seconds": 17628.223472098005, "env_steps_per_second": 4084.359386183286, "environment_steps": 72000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.5440809713610495, "eta_seconds": 1958.6914968997783, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.002569062088866758, "chaser_grad_norm": 0.31899478659033775, "chaser_new_fields": 62.54999923706055, "chaser_partner_vision": 0.984000027179718, "chaser_return": 10.590001106262207, "chaser_subspace_norm": 247.79714965820312, "collisions_per_episode": 7.800000190734863, "entropy": 0.13246314972639084, "explorer_grad_norm": 1.513938695192337, "explorer_new_fields": 71.3499984741211, "explorer_partner_vision": 0.984000027179718, "explorer_return": 49.525001525878906, "explorer_subspace_norm": 96.02532196044922, "final_distance": 1.7256439924240112, "max_param_abs": 2.4929258823394775, "policy_loss": -0.0048541350364530444, "value_loss": 2.3455803394317627}, "progress_fraction": 0.9, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T00:53:26.608918+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040006.6089342, "update": 18000} +{"device": "cuda", "elapsed_hours": 5.030915755465265, "elapsed_seconds": 18111.296719674952, "env_steps_per_second": 4085.8476974545474, "environment_steps": 74000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.4079120882809674, "eta_seconds": 1468.4835178114827, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_018000.safetensors", "metrics": {"approx_kl": 0.003319284843250614, "chaser_grad_norm": 1.6622390151023865, "chaser_new_fields": 60.60000228881836, "chaser_partner_vision": 0.971500039100647, "chaser_return": 10.744999885559082, "chaser_subspace_norm": 246.16616821289062, "collisions_per_episode": 8.40000057220459, "entropy": 0.14119615405797958, "explorer_grad_norm": 0.930143803358078, "explorer_new_fields": 67.5999984741211, "explorer_partner_vision": 0.971500039100647, "explorer_return": 44.53750228881836, "explorer_subspace_norm": 100.79618835449219, "final_distance": 1.716668725013733, "max_param_abs": 2.515129327774048, "policy_loss": -0.004607101465126107, "value_loss": 2.4525160789489746}, "progress_fraction": 0.925, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:01:29.682163+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040489.682175, "update": 18500} +{"device": "cuda", "elapsed_hours": 5.165626093922765, "elapsed_seconds": 18596.253938121954, "env_steps_per_second": 4086.8446006860286, "environment_steps": 76000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.27187505757488234, "eta_seconds": 978.7502072695764, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.004203352035322183, "chaser_grad_norm": 0.21013960242271423, "chaser_new_fields": 62.79999923706055, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.630000114440918, "chaser_subspace_norm": 285.9476623535156, "collisions_per_episode": 8.75, "entropy": 0.12066735327243805, "explorer_grad_norm": 0.754154309630394, "explorer_new_fields": 69.375, "explorer_partner_vision": 0.986750066280365, "explorer_return": 46.20000076293945, "explorer_subspace_norm": 102.30455017089844, "final_distance": 1.674743890762329, "max_param_abs": 2.5579209327697754, "policy_loss": -0.0044161222460274985, "value_loss": 2.2838812470436096}, "progress_fraction": 0.95, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:09:34.639383+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783040974.6393964, "update": 19000} +{"device": "cuda", "elapsed_hours": 5.302051708591656, "elapsed_seconds": 19087.38615092996, "env_steps_per_second": 4086.4683819580896, "environment_steps": 78000000, "error": null, "estimated_cost_usd": null, "eta_hours": 0.13595004381004244, "eta_seconds": 489.4201577161528, "healthy": true, "latest_checkpoint": "runs/mouse-run-run-2-transformer/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.003725360179970494, "chaser_grad_norm": 0.5451644361019135, "chaser_new_fields": 58.900001525878906, "chaser_partner_vision": 0.9655000567436218, "chaser_return": 10.727500915527344, "chaser_subspace_norm": 270.3601379394531, "collisions_per_episode": 8.675000190734863, "entropy": 0.12789610773324966, "explorer_grad_norm": 0.909948855638504, "explorer_new_fields": 68.0250015258789, "explorer_partner_vision": 0.9655000567436218, "explorer_return": 44.400001525878906, "explorer_subspace_norm": 111.15792846679688, "final_distance": 1.7669010162353516, "max_param_abs": 2.588141679763794, "policy_loss": -0.004116964178166427, "value_loss": 2.6845964193344116}, "progress_fraction": 0.975, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:17:45.771595+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041465.7716064, "update": 19500} +{"device": "cuda", "elapsed_hours": 5.40436246151166, "elapsed_seconds": 19455.704861441976, "env_steps_per_second": 4111.904480960076, "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-transformer/social/seed_0009/attempt_01/checkpoints/update_019000.safetensors", "metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "progress_fraction": 1.0, "schema_version": 2, "state": "running", "timestamp": "2026-07-03T01:23:54.090305+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041834.090314, "update": 20000} +{"device": "cuda", "elapsed_hours": 5.404370285466947, "elapsed_seconds": 19455.73302768101, "env_steps_per_second": 4111.898528119115, "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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", "metrics": {"approx_kl": 0.0012853747145520344, "chaser_grad_norm": 1.286025196313858, "chaser_new_fields": 57.650001525878906, "chaser_partner_vision": 0.986750066280365, "chaser_return": 11.644999504089355, "chaser_subspace_norm": 265.3517761230469, "collisions_per_episode": 9.75, "entropy": 0.11431185901165009, "explorer_grad_norm": 1.4314807802438736, "explorer_new_fields": 67.17500305175781, "explorer_partner_vision": 0.986750066280365, "explorer_return": 41.72500228881836, "explorer_subspace_norm": 115.21052551269531, "final_distance": 1.8594802618026733, "max_param_abs": 2.6002023220062256, "policy_loss": -0.002069776682645852, "value_loss": 2.937978982925415}, "progress_fraction": 1.0, "schema_version": 2, "state": "completed", "timestamp": "2026-07-03T01:23:54.118472+00:00", "total_environment_steps": 80000000, "total_updates": 20000, "unix_time": 1783041834.1184835, "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..233784655de726d72dba67375992fd941c17a8a3 --- /dev/null +++ b/social/seed_0009/attempt_01/status.json @@ -0,0 +1,40 @@ +{ + "device": "cuda", + "elapsed_hours": 5.404370285466947, + "elapsed_seconds": 19455.73302768101, + "env_steps_per_second": 4111.898528119115, + "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-transformer/social/seed_0009/attempt_01/checkpoints/latest.safetensors", + "metrics": { + "approx_kl": 0.0012853747145520344, + "chaser_grad_norm": 1.286025196313858, + "chaser_new_fields": 57.650001525878906, + "chaser_partner_vision": 0.986750066280365, + "chaser_return": 11.644999504089355, + "chaser_subspace_norm": 265.3517761230469, + "collisions_per_episode": 9.75, + "entropy": 0.11431185901165009, + "explorer_grad_norm": 1.4314807802438736, + "explorer_new_fields": 67.17500305175781, + "explorer_partner_vision": 0.986750066280365, + "explorer_return": 41.72500228881836, + "explorer_subspace_norm": 115.21052551269531, + "final_distance": 1.8594802618026733, + "max_param_abs": 2.6002023220062256, + "policy_loss": -0.002069776682645852, + "value_loss": 2.937978982925415 + }, + "progress_fraction": 1.0, + "schema_version": 2, + "state": "completed", + "timestamp": "2026-07-03T01:23:54.118472+00:00", + "total_environment_steps": 80000000, + "total_updates": 20000, + "unix_time": 1783041834.1184835, + "update": 20000 +} diff --git a/triton_equivalence.json b/triton_equivalence.json new file mode 100644 index 0000000000000000000000000000000000000000..25700886c92dd971ac253eacbfec026c21606a1a --- /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-transformer/triton_equivalence.json", + "completed_at": "2026-07-02T14:31:48.654521+00:00", + "device": "cuda", + "grid_size": 10, + "provenance": { + "command": "scripts/verify_triton_env.py --output runs/mouse-run-run-2-transformer/triton_equivalence.json", + "created_at": "2026-07-02T14:31:48.654546+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": "2026-07-02T14:31:40.454066+00:00", + "status": "ok", + "steps": 100, + "vision_radius": 3 +}